education

Education


Round Corner
Department of Computer and Information Science

Forberedende- og fordypningsprosjekt 2021

Marker valg for å avgrense hvilke oppgaver som skal vises.

Studieretninger

Datateknologi (865)






Informatikk (635)





Helseinformatikk (25)


Faglærere (65)

































































Sorter etter:

Oppgaveforslag (382)

AI & eHealth: Case Base Evolution

A interesting problem in personalized decision-support systems for improving an individual patient’s health, is to combine general clinical guidelines with past experiences of that same or similar patients. In the EU project selfBACK, in which we tightly cooperate with the Department of public health and nursing at NTNU, we combine rule-based reasoning with case-based reasoning to capture these two knowledge types. The target problem is Low-Back Pain, and the aim of the project is to develop and thoroughly test a mobile phone app that will give a patient advice on how to improve his/her lower back conditions, in a short or long term. Activity data is continuously captured via a wrist band and used alongside subjective information on pain and functionality. This project will focus oncase-based reasoning and data analytics investigating the evolution of cases over time.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Human Activity Recognition from Accelerometer Data in HUNT4

Data captured by body worn sensors provides an excellent opportunity for assessing the physical activity of patients and hence creating behavioral profiles over time. Particularly patients with chronic disease can receive tailored advice on how to increase their activity and hence improve their overall life quality.

Status: Tildelt     Egnet for: En student     Lenke: plink

Machine Learning to improve the Air Quality in Trondheim

Over the last year we have created a dataset with information on air quality data and we started to explore machine learning methods for predicting the air quality for the next 12/24/48 hours as well as visualize the results.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Recognition of Sleep Patterns on Sensor Data Streams (HUNT4)

Data captured by body worn sensors provides an excellent opportunity for assessing the physical activity of patients and hence creating behavioral profiles over time. Particularly patients with chronic disease can receive tailored advice on how to increase their activity and hence improve their overall life quality.
The focus and challenge for this project and master thesis is the selection, implementation and improvement of pattern recognition and data mining techniques to identify sleep patterns from sensor data. The data will be provided by NTNUs medical faculty (DMF), while this thesis should focus on the data analysis. The captured data sets will be streaming data from two acceleration sensors recorded at 100 Hz.
During this work you will build on previous work that does a basic classification of awake/sleep phases and extend the model finding various sleep patterns. Also investigating different sleep stages is a possibility. The implementation will be evaluated in collaboration with DMF, who is also providing background information on the data.
Once the experimental set up is created, different existing algorithms should be evaluated and their strength and weaknesses pointed out. Based on this analysis, a follow-up master thesis can be defined focusing on improving existing algorithms and validated in a real world setting.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Reinforcement learning to actuate in Trondheim's air quality

Under the AI4EU project (https://www.ai4eu.eu) a urban simulator of Trondheim is being developed, based on the SUMO traffic simulator. SUMO is an open source traffic simulator that is developed to simulate realistic road networks. As a side effect it also models the pollutant emissions of vehicles which allows us to model the effect of traffic pollution in an urban scenario. Traffic data is publicly available which can be inputted to the traffic simulator and obtain realistic simulated traffic patterns. This simulator serves as a tool that captures the realistic patterns of air quality data and can be used as an environment to train autonomous agents, both in what concerns to improve air quality levels and the quality of the information on pollution levels.
Therefore, the goal of this thesis is the deployment of agents that learn, through reinforcement learning algorithms, from data coming from this simulator. The goal of the agent can be defined on the setup of the project. Some examples are:

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Towards a realistic urban simulator of Trondheim

Under the AI4EU project (https://www.ai4eu.eu) a urban simulator of Trondheim is being developed, based on the SUMO traffic simulator. SUMO is an open source traffic simulator that is developed to simulate realistic road networks. As a side effect it also models the pollutant emissions of vehicles which allows us to model the effect of traffic pollution in an urban scenario. Traffic data is publicly available which can be inputted to the traffic simulator and obtain realistic simulated traffic patterns. The final goal is to have a tool that captures the realistic patterns of air quality data and can be used as a tool either for evaluation of future scenarios in decision support systems and as an environment to train autonomous agents, both in what concerns to improve air quality levels and the quality of the information on pollution levels.
Therefore, the goal of this thesis is to build up from the current simulator development and improving its realism, including but not limited to:
• Inclusion of more pollution sources: currently, only emissions from traffic are modeled (which contributes mostly to NOx levels) but air quality levels are influenced also by ships, wood burning fireplaces, air dust (which depends on the season), etc.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Intent classification for conversational AI

Conversational agents, such as Apple’s SIRI, Amazon’s Alexa, or the Google Assistant are capable of handling a broad range of requests. In the context of such conversational AI systems, the objective of this project is to develop a machine learned classifier that determines the underlying intent of user utterances.
The project involves the following specific tasks:

Status: Tildelt     Egnet for: En student     Lenke: plink

Narrative-driven recommendations

Many online services provide users with recommendations, to help them find items of interest in the enormous space of available choices. Popular examples of such recommender services include videos (YouTube), music (Spotify), movies (Netflix), online shopping (Amazon), etc. These services typically base their recommendations on items liked/disliked by users.

Status: Tildelt     Egnet for: En student     Lenke: plink

Preference elicitation in conversational recommendations

Recommender systems have become indispensable tools that help users navigate large collections to find items of potential interest. Popular examples of such recommender services include videos (YouTube), music (Spotify), movies (Netflix), online shopping (Amazon), etc. These services typically base their recommendations on items liked/disliked by users.

Status: Valgbart     Egnet for: En student     Lenke: plink

Ranking items based on soft attributes

Recommender systems provide personalized recommendations of products or services to users, based on their preferences. A key element of modern conversational recommender systems is the ability for users to provide feedback on their likes and dislikes. Ideally, such systems should be able to incorporate any natural language user feedback on items.

Status: Valgbart     Egnet for: En student     Lenke: plink

Scientific literature recommendation

Many online services provide users with recommendations, to help them find items of interest in the enormous space of available choices. Popular examples of such recommender services include videos (YouTube), music (Spotify), movies (Netflix), online shopping (Amazon), etc. In this project, we wish to provide recommendations for scientific literature to researchers.

Status: Tildelt     Egnet for: En student     Lenke: plink

Database opportunities at Huawei Research lab, Trondheim

Huawei has established a database research lab in Trondheim.
They have different projects available for students.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Learned index and updates

There is a lot of work going on in learned indexes. The project should survey how updates are handled in learned index and suggest improvements to tackle a high load of inserts and updates.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Managing big data volumes

Oppgaven går ut på å studere metoder for parallell innsetting av store mengder data. Oppgaven er formulert generelt, men kan ha spesifikk fokus på data fra visse anvendelser, f.eks. geografiske data fra satelitter.

Status: Valgbart     Egnet for: En student     Lenke: plink

Merging of geodata

Merging of data from space-filling curves used in e.g. R-trees.
In the era of velocity of big data, geodata needs to be merged
just like key-based data is merged in LSM-trees to cater for high
insertion rates. Investigate methods for doing this. Merging in LSM trees is also a hot research topic.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Merging of offline databases

The problem is to find solutions to merging of databases which have been offline for a while. This may for instance be rescue teams working in areas with no connection, and which needs to synchronize its work.

There are several possible  solutions to this

- CRDTs

- Mergeable datastructures

- Operational transformations

- .....

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

MySQL-related problems

There are multiple opportunities for students interested in MySQL technology.
Take contact to talk about possibilities.
This is done in cooperation with Oracle's MySQL team.

Status: Valgbart     Egnet for: En student     Lenke: plink

NoSQL/NewSQL databases

Flere muligheter her. Ta kontakt.

Status: Valgbart     Egnet for: En student     Lenke: plink

Creativity as a Quality Factor in Software Development Processes

Current practices in the software development processdo not include specific needs to ensure the creativity as a quality aspect. This research aims to investigate creativity as a quality factor in the software development process.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Documentation of Security in Agile Projects

Security breaches are happening all around us. Software systems have developed to the point that we use and depend upon them daily in the same way that we depend upon traditional infrastructures and utilities. The value of sensitive information in computer systems is constantly increasing, and the same can be said for the corresponding threats, but measures to reduce the resulting vulnerabilities are not developed at the same pace.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Quality Assurance for AI-based Systems

The purpose of this thesis, that can be chosen as project and as thesis, by one or two students and also by several groups is to support the implementation of AI software inside the Software Development process. Each student (or couple of students) will study a sub theme of the theme SE and AI, study the literature, plan an empirical investigation to collect data from processes that involve software engineers and software companies.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Security Risk Management in Agile Software Development

Security breaches are happening all around us. Software systems have developed to the point that we use and depend upon them daily in the same way that we depend upon traditional infrastructures and utilities. The value of sensitive information in computer systems is constantly increasing, and the same can be said for the corresponding threats, but measures to reduce the resulting vulnerabilities are not developed at the same pace.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Software Engineering and AI

Software engineering and AI

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Software security for Software for Children

Software security for Software for Children

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Software Security in Agile

Security breaches are happening all around us. Software systems have developed to the point that we use and depend upon them daily in the same way that we depend upon traditional infrastructures and utilities. The value of sensitive information in computer systems is constantly increasing, and the same can be said for the corresponding threats, but measures to reduce the resulting vulnerabilities are not developed at the same pace.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Software Security in Agile (Management of Security Technical Debt)

Security breaches are happening all around us. Software systems have developed to the point that we use and depend upon them daily in the same way that we depend upon traditional infrastructures and utilities. The value of sensitive information in computer systems is constantly increasing, and the same can be said for the corresponding threats, but measures to reduce the resulting vulnerabilities are not developed at the same pace.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Software Security Requirements Elicitation and Testing in Agile/DevOps

Security breaches are happening all around us. Software systems have developed to the point that we use and depend upon them daily in the same way that we depend upon traditional infrastructures and utilities. The value of sensitive information in computer systems is constantly increasing, and the same can be said for the corresponding threats, but measures to reduce the resulting vulnerabilities are not developed at the same pace.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

The Role of Security Champions in Agile Teams

Security breaches are happening all around us. Software systems have developed to the point that we use and depend upon them daily in the same way that we depend upon traditional infrastructures and utilities. The value of sensitive information in computer systems is constantly increasing, and the same can be said for the corresponding threats, but measures to reduce the resulting vulnerabilities are not developed at the same pace.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Digital tjenesteinnovasjon for trygge barn i tannbehandling

Tannbehandlingsangst regnes å være den viktigste årsaken til at pasienter ikke møter til tannbehandling, og kan føre til alvorlige svekkelser i oral helse, redusert livskvalitet og i tillegg betydelige utgifter for den offentlige tannhelsetjenesten. Prevalensberegninger av tannbehandlingsangst hos barn og ungdom viser opp mot 20 %, og barn med tannbehandlingsangst har dårligere tannhelse og uteblir oftere fra tannbehandling sammenlignet med andre barn.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

ExAct: Tangible Interactive Technology for Physical Rehabilitation and Therapy

The advancements in interactive technology over the last decades have given rise to a variety of movement-based applications for health purposes such as in rehabilitation and physical therapy. Despite the potential of such applications, this form of technology is still far from being part of the standard equipment in physiotherapists’ toolbox, ready at hand in their everyday work with patients.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Large-scale agile software develpment

Agile software development methods have gained increasing popularity in software development projecs. Agile methods prescribe practices for development, and were first used in small projects with little criticality. However, such methods are increasingly used in large projects, and this project will investigate how the practices are adapted and combined with traditional practices to function effectively in large scale. A first generation of large-scale agile methods combined advice from methods such as Scrum with advice from project management. A second generation of methods are currently taken up by the global software industry, with methods such as the Scaled Agile Framework, Large-Scale Scrum, the Spotify model and Disciplined Agile Delivery. 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Smidig transformasjon

Smidige metoder oppstod i team som jobbet med systemutvikling, men har blitt tatt i bruk av større deler av organisasjoner og ført til endringer i ledelse av prosjekter og større programmer. Smidige metoder legger vekt på beslutningstaking på lavt nivå i selvstyrte team, tett kontakt med kunde og fleksibilitet i arbeidsprosesser for å kunne håndtere endringer underveis i utviklingsprosjekt. Når organisasjoner ønsker å ta i bruk denne type metoder i større deler av organisasjonen støter de ofte på utfordringer med eksisterende hierarki, at det kan være vanskelig å bli enige om arbeidsprosesser med miljø med andre oppgaver enn programvareutvikling og generalt at det er motstand mot organisatoriske endringer. Denne oppgaven vil bestå i å først gjøre et litteraturstudie på temaet “smidig transformasjon” innen fagområdene systemutvikling og informasjonssystemer, og koble med relevante andre felt som for eksempel endringsledelse. En videreføring i masteroppgave vil kunne innebære et empirisk studie av en eller flere organisasjoner som gjennomfører en “smidig transformasjon”.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Teamwork in software development

Teamwork is central in software development, and is currently a topic much addressed in agile software development where the development is performed in small, self-organized teams. Improving the efficiency and effectiveness in software development will therefore often involve improving the way the teamwork is organized. Several team performance models have been suggested in the research literature, and there is a growing number of empirical studies of teamwork in software development with focus on specific characteristics for agile development teams and distributed or virtual teams.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

DIGVID- Digital competencies for teaching

DIGIVID is a newly started European project involving Austria, Germany, and Austria. In the DIGIVID project, we focus on the education of lecturers, in-service teachers and teacher-students to educate them to become information literate and digital-savvy to ensure high-level teaching and learning quality in online teaching and learning environments.
TThe project aims at developing DIGIVIDget, a widget that presents the DIGIVID online teaching and learning curriculum based upon the European Digital Competence Framework for Citizens (DigComp 2.1) and the European Digital Competence Framework for Educators (DigCompEdu). DIGIVID empowers lecturers, teachers and teacher-students to learn time and place-independent on their mobile devices or on (desktop-)PCs to prepare them for the demands of today’s information society and working contexts.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Increasing awareness about privacy and personal data with games

Advancements in information technology have made people less aware of the collection and usage of personal data. As a result, individuals rarely have clear knowledge of what information other people and firms store about them or how that information is used. The problem is made even more pressing with the increasing adoption of IoT and interactive objects, promoting new forms of interaction and data collection for which new strategies needs to be developed.
This task aims to investigate how to use serious games and scenario tools to evoke reflection about sharing of personal data and privacy and promote learning about these issues. Games in this context are intended as a way to help players to see things differently and reflect on their actions, their consequences, and tradeoffs of one´s choices.
The task might be specialized to consider (i) challenges to personal data set by different types of technology (e.g. mobile, IoT, social media, …), (ii) different types of games (e.g. mobile games, board games, …), (iii) different type of users (e.g. children, the elderly, …). These specialization will be done at the beginning of the semester based on the interest and competencies of the student.
The task will start with the identification of some relevant scenarios and a study of current literature. It will then continue with the iterative development of a prototype to be evaluated with users.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Raising Awareness of Climate Change with Virtual, Augmented and Extended Reality

Virtual, Augmented and Extended Reality (VR/AR/XR) are related technologies that can provide engaging environments for learning, enable 3D visualizations of complex concepts and allow experiencing potentially dangerous situations in safe settings. VR is also called “the ultimate empathy machine” and may provoke strong emotional responses among users.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Running and Learning in Virtual Reality: Active Educational Games using VR Treadmill Virtuix Omni

Virtual Reality (VR) opens new possibilities in education, as it provides engaging and immersive learning environments, enables 3D visualizations of complex concepts and allows training on potentially dangerous situations in safe settings. Past and ongoing master projects at NTNU resulted in several working VR prototypes, especially for teaching math, climate change, historical reconstructions and medical training (see e.g. https://gemini.no/2019/02/undervisning-pa-mount-everest-og-mars/)
Active physical engagement (walking or running) with gaming elements (e.g. first-person-shooter) might contribute to a more engaging and immersive learning experience. This master thesis will focus on design principles and tools for developing active educational games for VR treadmill Virtuix Omni https://www.virtuix.com/. A working prototype for Virtuix Omni for such a game where users learn about sea level rise by running around flooded virtual Trondheim and ‘shooting’ pollution sources has been developed by students at IMTEL lab and successfully presented at Big Challenge festival June in 2019: https://drive.google.com/file/d/19csS9UmaGkaGJO_jizOC3wt-jIQbSyit/view?usp=sharing. Another example could be historical quests or sports education. One of research questions could be to investigate whether such active approach contributes to better learning experience.
The students will have access to a very well-equipped IMTEL VR lab (https://www.ntnu.edu/ipl/imtel) containing Valve Index, HTC Vive/Vives Pros, Vive Cosmos, 2 Magic Leaps, several Hololenses 1 and 2, Mixed Reality headsets, Oculus Quests, Oculus Rifts, VR treadmill Virtuix Omni, VR laptops etc. A significant number of the VR/AR equipment is portable and can be used at home shall the pandemic situation and campus closure be repeated.


Supervisors: Monica Divitini, Ekaterina Prasolova-Førland ekaterip@ntnu.no (IPL, NTNU)
 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Supporting teachers of programming in secondary schools

Teachers with the responsibility to teach programming in secondary schools are facing a number of challenges: the subject is changing fast, the curriculum is underspecified, ...

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

TILES - a toolkit for the Internet of Things

Tiles is a toolkit to facilitate the design of innovative solutions for societal challenges that meet the UN's Sustainable Development Goals. It empowers groups of non-experts in designing physical objects augmented with sensors and actuators to provide computer interactivity and connectivity (Internet of Things). More information about the toolkit is available at https://www.tilestoolkit.io/

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Tools for teaching and learning Computer Science

This task focuses on the study of tools that promote teaching and learning in Computer Science at the university level. The task is explorative in nature and will address questions like: which tools are used in the context of CS education? Are tools used in industry used in education? and how? How do specific and generic learning tools co-existing? 

Status: Valgbart     Egnet for: En student     Lenke: plink

Virtual, Augmented and Extended Reality for Remote Collaborative Learning in the context of COVID-19 crisis (in collaboration with Kavli Institute for Systems Neuroscience)

Virtual, Augmented and Extended Reality (VR/AR/XR) can provide rich and interactive experiences for remote distributed learners, which is highly relevant in the current context of COVID-19 epidemic. The traditional solutions such as Zoom often lack the sense of presence, rich interactivity and realistic experiences that are essential for many learning situations (such as lab work) which can be supported in VR/AR. Delays in vaccinations and recurrent outbreaks make it necessary finding new methods and strategies to facilitate rich remote learning.
The goal of this master project is to perform research on remote interactive educational tools and methods in VR/AR and contribute to NTNU preparedness in the context of pandemic. One of the possibilities is joining the on-going collaboration with the Nobel prize winning team at Kavli Institute for Systems Neuroscience, where a solution for remote collaborative brain dissection and anatomy teaching (with Hololens 2) is currently in progress: https://www.youtube.com/watch?v=QyZNPSCvkb4

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Virtual, Augmented and Extended Reality for virtual field trips, virtual labs and language learning

‘Traditional’ teleconferencing and online learning solutions such as Blackboard and Zoom often lack the sense of presence, rich interactivity and realistic experiences that are essential for many learning situations such as lab work, practice, field trips and foreign language excursions. Virtual, Augmented and Extended Reality (VR/AR/XR) technologies can provide engaging collaborative environments for learning, enable 3D visualizations of complex concepts and allow exploration of situations not possible otherwise, especially in the context of the pandemic.
Depending on the interests of the student(s), this master thesis will focus on design principles and tools for developing educational VR/AR experiences for one of topics: 1) virtual field trips in collaboration with VR-Learn project and Department of Geography (see e.g. https://youtu.be/24lFLQUjuew); 2) virtual labs in collaboration with NTNU Labforum; 3) VR language learning in collaboration with EU DC4LT project (https://www.dc4lt.eu/immersive-technologies-for-language-learning/ ). For examples of earlier similar projects created by our master students, see e.g. https://gemini.no/2019/02/undervisning-pa-mount-everest-og-mars/.
The students will have access to a very well-equipped IMTEL VR lab (https://www.ntnu.edu/ipl/imtel) containing Valve Index, HTC Vive/Vives Pros, Vive Cosmos, 2 Magic Leaps, several Hololenses 1 and 2, Mixed Reality headsets, Oculus Quests, Oculus Rifts, VR treadmill Virtuix Omni, VR laptops etc. A significant number of the VR/AR equipment is portable and can be used at home shall the pandemic situation and campus closure be repeated.


Supervisors: Monica Divitini, Ekaterina Prasolova-Førland ekaterip@ntnu.no (IPL, NTNU)
 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Workplace training in Virtual, Augmented and Extended Reality in collaboration with NAV and local industry

The unemployment in Norway and worldwide has increased dramatically as a result of the COVID-19 pandemic, raising the need for developing innovative methods for workplace training and career guidance. In this project we will investigate how the use of Virtual, Augmented and Extended Reality (VR/AR/XR) technologies and gaming elements can 1) motivate and inform young job seekers on their way to work and 2) contribute to faster skill acquisition for new employees. The virtual experience will allow the users to train in unfamiliar situations (job interview, typical tasks at different workplaces, interaction with others) in a safe setting, thus mastering the corresponding real world situation. Through the simulation of a workplace or an industry (e.g. aquaculture, car workshop or pharmacy), the job seekers can immerse into different workplaces and try out typical tasks, for example, salmon feeding or administering a COVID-19 vaccine.
This master thesis will focus on design principles and tools for developing and evaluating workplace training in VR/AR/XR, in connection with the ongoing project financed by NAV (Norwegian Labour and Welfare Administration), https://www.ntnu.edu/imtel/virtual-internship. The project has so far resulted in several prototypes for workplace training and job interview training in VR and received international recognition (e.g. Best Demo Award at EuroVR 2018 and Breakthrough Auggie Award finalist) and broad media coverage https://memu.no/artikler/gir-ungdom-en-virtuell-jobbsmak/. The student(s) will work in close collaboration with NAV, local industries and our European partners (e.g. Performance Augmentation Lab in Oxford). We are collaborating with several research environments at NTNU and European VR/MR labs on this and other projects.
The students will have access to a very well-equipped IMTEL VR lab (https://www.ntnu.edu/ipl/imtel) containing Valve Index, HTC Vive/Vives Pros, Vive Cosmos, 2 Magic Leaps, several Hololenses 1 and 2, Mixed Reality headsets, Oculus Quests, Oculus Rifts, VR treadmill Virtuix Omni, VR laptops etc. A significant number of the VR/AR equipment is portable and can be used at home shall the pandemic situation and campus closure be repeated.


Supervisors: Monica Divitini, Ekaterina Prasolova-Førland (ekaterip@ntnu.no) & Mikhail Fominykh
 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Biologically-Plausible Neural Networks

In this open-ended project, students will investigate a wide range of learning algorithms for neural networks, all of which are much more biologically realistic than the most popular (and successful) tools used in Deep Learning.  These more biological approaches employ "local" learning rules similar to the famous "Hebb Rule" of neuroscience, whereas contemporary DL algorithms update their weights based on very "global" forms of information, i.e. the effect of a weight in layer 3 on the output at layer 62.  These local learning approaches have existed since the beginnings of neural-network research, and one of them in particular, the Restricted Boltzmann Machine (RBM) was the breakthrough that ushered in the current DL revolution.  Other, more recent, work shows that local approaches can achieve similar results to classic backpropagation, as shown in these papers:

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Birdsong classification using Deep Learning

To assist scientists in remotely identifying wildlife, deep neural networks can be used to localize and classify visual and/or auditory characteristics of various species, thus saving humans from online monitoring and/or poring through hours of recordings.  This project involves identifying birdsongs in complex auditory landscapes (i.e. recordings with assorted background noise in natural settings).  Following the same basic task description as the 2020 Kaggle competition for birdsong recognition:

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Deep Learning for Pollen Dating

Anthropologists, archaeologists and climatologists use many tools for dating and reconstructing environments thousands and millions of years in the past.  One such tool, palynology (pollen dating) uses microscopic pollen samples, consisting of hundreds or thousands of diverse pollen grains, to identify characteristic "pollen zones", which often provide unique time tags.  Unfortunately, this requires considerable human labor, as scientists stare for hours at 3-dimensional microscopic images full of pollen grains of many different species.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Deep Reinforcement Learning (DRL) for International Diplomacy

In 2020, Google Deepmind published a paper on the use of DRL for playing a multi-player (7) game based on the diplomatic (or lack thereof) interactions between nations.  The game requires simulated governments to compete AND COOPERATE to achieve their goals.  A copy of Google Deepmind's paper can be found here:

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Deep Reinforcement Learning for Robotic Manipulation

This project involves the use of Deep Reinforcement Learning (DRL) to train a robotic arm to pick up compliant (i.e. soft, deformable) objects.  This project is lead by Sintef OCEAN.  For more information, please read the following detailed project description:

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Explorations of Evolutionary Computation for Network Design and Control

 Evolutionary Computation (EC) facilitates a very creative, open-ended, automated design process that has led to a wide variety of inventions, from satellite antennas to furniture to robots.  In this project, the general target area is networks, and will be used to determine both the overall topology (i.e. interconnectivity patterns) and the individual behaviors of the network nodes.  The project itself is "open ended" in that a student may suggest an application area where networks are a natural modelling tool, such as power or communication grids, or even the spread of disease (where networks are often used to assess pandemic risks).

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Network-Based Pandemic Modelling

The spread of pathogens involves a complex combination of factors such as incubation time, virulence, and contagiosity of the virus; behavioral tendencies and demographics of the carriers (a.k.a. vectors); and the general composition and connectivity of the environment.  An additional factor is the decisions by individuals to vaccinate themselves (or not).  In this (rather open-ended project), students will implement and deploy tools for building abstract graph-theoretic models of environments (thus permitting wide variance in the capacity and connectivity of regions) within which simulated pathogens and vectors interact.  Via thorough experimentation, interesting conclusions should be drawn concerning relationships between environmental topology, important viral and vectorial characteristics, and human choice.  No background in biology is required, but interested students must be willing to read extensively on relevant topics within epidemiology and network theory, a well-explored combination in the scientific literature.  The use of techniques from bio-inspired artificial intelligence will also be encouraged.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Autotuning and/or ML for HPC -- several potential projects outlined

The following subprojects may be refined/altered to suite the background and interests of the student selecting this project.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Communication avoiding Conjugate Gradient algorithm programming on GPUs using CUDA

In this project you will analyze, implement and test a parallel Conjugate Gradient algorithm on newer CUDA GPUs.

Status: Valgbart     Egnet for: En student     Lenke: plink

Exciting topics in HPC, Parallel Computing, and Cloud Technologies

If you have a clear idea of an exciting  topic you want to work on in HPC (high-performance computing), including Parallel Computing and Cloud Technologies, contact me (Anne Elster elster@ntnu.no)  to see if we are a fit and how feasible you project is as a fall project and/or Master thesis.

 

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Improving The HPC-Lab GPU-based b Snow Simulator

Throughout the last decade, several master students have been working on a real time snow simulator at NTNU’s HPC-lab in Trondheim. The purpose of the simulator is to predict and visualize how snow will cover a landscape over time and the possibility for avalanches, based on different wind and terrain formations.

Status: Valgbart     Egnet for: En student     Lenke: plink

Investigating New GPU Features for Performance (NVIDIA Tegra and/or Betsy)

Look into how effective the current optimization techniques are for GPUs on the newest platforms, including  exploring how to use these GPU´s tensor processors for HPC applications and/or how selected benchmarks scale on the new Supercomputer Betsy  at NTNU.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Master theses topics in collaboration with CERN Openlab

These theses topics will be defined in collaboration with CERN for those students interested their Master program. Students participating in the CERN OpenLab summer program are particularly encouraged to apply.

Status: Valgbart     Egnet for: En student     Lenke: plink

Optimizing 4D CT Computation for Performance Through GPU Computing and AI

This project will be done in collaboration with Prof Dag Breiby´s research group at NTNU Physics, an is part of our joint NFR FRIPRO project on computational microscopy.

Status: Valgbart     Egnet for: En student     Lenke: plink

Seismic Simulations on GPU - Fall 2021

This project is a collaborations between HPC-Lab at IDI and Prof. Børge Arntsen, Petroleum Engineering and /or Schlumberger.

Details to be flushed out shorly, but feel free to email Elster (elster@ntnu), if interested.

 TDT4200 or equivalent required. I

Status: Valgbart     Egnet for: En student     Lenke: plink

A RISC V Oberon compiler backend

Current software systems require multiple gigabytes of memory and multicore systems to run efficiently. However, the advances in hardware speed and capabilities due to the advancement of Moore's Law that enabled this so-called "software bloat" are coming to an end. Accordingly, a movement advocating for lean software has surfaced [1]. This is based on the observations that earlier computers were able to provide a useful working environment including a graphical user interface and networking with far fewer resources.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Creating a microkernel-based multi-server operating system

Background

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Program slicing for intermittent computing in IoT systems

Many systems in the Internet of Things (IoT) run on energy harvesting, e.g. using solar cells, and thus may suffer from power loss during operation. Intermittent computing [1] is the general name of a research area that develops and implements methods to enable software to run on systems with interrupted power supply without having to restart.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Self determination, privacy and sustainability in IoT systems

This is an umbrella project that can result in of a number of subtopics for specific projects. Accordingly, the specific topics will be defined in details together with the interested student(s). Please get in touch with me for further information and/or discussing details.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Software defined memory

Non-functional properties of computer systems, such as runtime and energy consumption. are becoming increasingly unpredictable due to the introduction of memory hierarchy levels and, especially, complex caches.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Software infrastructure for a joint robotics project

Topic

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Design and evaluation of digital platforms

Please contact me before choosing this task.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Digital platforms in the public sector (including NAV and Trondheim municipality)

Please contact me before you select this task as your first priority.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Digital transformation and software engineering

Please contact me before selecting this task.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Empirical investigation of collective intelligence tools used in the public sector

This is a task connected to an R&D project called COLDIGIT at SINTEF. A researcher from SINTEF will be co-supervisor. The task will require English as working language and the candidates will be working in the context of COLDIGIT.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Technologies for digital empowerment and digital coping

Please contact me before selecting this task.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

An emotion sensitive chatbot for improved mental health

Chatbots that can carry on a dialogue with no direct goal (chit-chat) have developed rapidly in recent years and are often based on deep reinforcement learning or transfer learning strategies, where deep learning is used to pre-train models on large out-of-domain datasets and then fine-tune on some in-domain data. This project would look at dialogues related to mental health and predict the user's mood: The language that people use and the ways they use it can provide information about their mental health state, with several researchers over the last 50 years having based mental health predictions on the users' choice of words and on word frequencies. The results clearly indicate that increased usage of emotion-oriented language is related to increased mental well-being, with the people benefiting the most from talking, writing and chatting about their mental health being those that also initially use a higher frequency of words conveying positive emotions.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Automatic detection of pro-eating disorder (pro-ED) social media users

Pro-eating disorder groups (pro-ED) are social media sub-cultures that encourage disordered and dangerous eating behaviours, e.g., Pro-Ana (pro-anorexia), Pro-Mia (pro-bulimia) and Thinspro (Thinspiration, a combination of “thin” and “inspiration”). Automatic detection of users sharing, supporting or following pro-ED content can provide information for understanding and preventing eating disorders, as well as for social media moderation. Data on some such users on Twitter have already been annotated, but to fully apply machine learning algorithms such as deep learning to the problem, more data need to be gathered, tentatively from various sites such as Twitter, Reddit and Tumblr. The thesis work would then experiment with applying various machine learners to this data.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Automatic music transcription with deep learning

The field of Music Information Retrieval (MIR) has gained a lot of momentum in the last couple of years with the advancement of deep learning methods, while music source separation has become feasible as a pre-processing step following several recent projects. Polyphonic piano music transcription has also seen progress with the Onsets and Frames model (Hawthorne et al. 2018) based on bi-directional Long Short-Term Memory recurrent neural networks. Fully automatic music transcription would be highly valuable for musicians not only as an educational tool but also when arranging music and combining acoustic and electric music in live performances. Data collections such as the Million Song Dataset and derivatives from it like the Lakh MIDI Dataset and the Lakh Pianoroll Dataset gives possibilities for large scale supervised learning in this domain.

Status: Valgbart     Egnet for: En student     Lenke: plink

Code-switching and multi-lingualism in social media

When two individuals who are bi- or multi-lingual in an overlapping set of languages communicate, they tend to switch seemlessly and effortlessly between the languages (codes) they share. Such code-switching is most prominent in spoken language conversations, but also occurs frequently in social media texts that are fairly informal and conversational in nature. The aim of this project is to apply various machine learning methods to such code-switched texts from Twitter, Facebook or Whatsapp, and to, e.g., identify the language of each word or to annotate the texts with part-of-speech tags or utterance boundaries.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Computational Creativity

To be creative, we need to produce something which is new, meaningful and has some sort of value. Computers are able to support humans in creative processes, but to also themselves be creative or to assess if an idea or a product is creative. A master thesis project on computational creativity can investigate any creative field matching the interests and backgrounds of the student or students (language, design, music, art, mathematics, computer programming, etc.), and concentrate on one or several aspects of computational creativity, such as the production, understanding or evaluation of creativity, or on computer systems that support human creativity.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Computational Creativity and Art

Creativity can be found in nature and in humans, but also in computers, and entails to produce something which is new. However, just “newness” isn’t a sufficient condition for us to consider an idea to be creative, it also has to have some value and meaning: If a 2 year old draws some lines on a paper, we rarely consider it to be art; while if a grown-up does the same, we interpret it as having some deeper meaning – and if the grown-up signs the paper with a well-known artist name, we attribute both an underlying meaning and a monetary value to it. Creativity is thus something which isn’t only a result of the effort of a producer, but also very much the result of how the result is viewed by the consumer.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Computational Creativity with Reinforcement Learning

There are three basic machine learning paradigms: supervised learning, unsupervised learning, and reinforcement learning. Supervised learning requires data instances annotated with the correct output labels, but the other paradigms assume no such labelled data. Reinforcement learning is primarily concerned with what actions an intelligent agent (or group of agents) should take in a specified environment in order to maximise some cumulative reward, while simultaneously exploring the environment and exploiting previously accumulated knowledge. The environment is commonly formulated as a Markov decision process or as a partially observable Markov decision process (POMDP), depending on whether the agent can observe the current environmental state directly or whether the observed states are distorted by noise.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Computational Lingustic Creativity

Computational linguistic creativity can be aimed at creating systems that either are creative themselves (e.g., generate poetry, write lyrics to music, produce analogies or metaphors; or chatterbots), or try to understand creativity (e.g., identify sarcasm, understand humour or interpret rhymes), or support humans in creative processes (such as PhotoShop in the image domain), or evaluate creativity.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Computational Musical Creativity

Computers have been used in music both as support for creativity and as creative agents themselves, and both for the composition of the music scores and for writing lyrics. The first algorithmic composition system appeared already in the 1950s (the Illiac suite, Hiller & Isaacson 1958), and since then rule-based systems, stochastic methods, grammar-based methods, neural networks, and evolutionary methods have all been utilised to compose music, and/or for generating lyrics. A master thesis on the topic could address any of these strands and approaches, depending on the student(s) background and interests.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Emphasis selection for short texts in visual media

Automatic captioning of images and videos involves producing short texts matching the visual content. To fully convey the message by the visual media, parts of the texts may need to be emphasised, which currently is mainly a manual task. Some applications utilise automatic methods, but commonly focus on visual attributes, such as word length. However, to convey an idea efficiently, the attributes taken into account should rather be semantic, concentrating on the key parts of the texts as meant by the user, but without relying on information about the user intent being explicitly provided. The project would explore the combination of images/videos with very short (fewer than 10 words) texts, with an aim to rank and select the best candidates when multiple emphasis patterns are possible.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Entity-level sentiment impact analysis in social media texts

A key aspect of sentiment analysis is identifying the target(s) of the opinion, that is, to determine which entities in a text the expressed sentiment relates to. Exploring how textual entities are related to a text’s overall sentiment can yield information on how given entities are portrayed in social media, e.g., on Twitter. This requires the application of sentiment analysis techniques as well as named entity recognition and linking, and the use of heuristic or grammatical features to determine entity relevance and sentiment strength.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Evolutionary Algorithms for Language Processing

Natural language processing grapples with an ever-changing and moving target. The focus of study, natural language, is natural because it changes, interacts and evolves in various directions. The bio-inspired computational methods described as evolutionary computation and/or genetic algorithms create computational models that evolve a population of individuals to find a solution to a given problem. This project will investigate how evolutionary computation can been employed in some natural language processing task, ranging from efforts to induce grammars to models of language.

Status: Valgbart     Egnet for: En student     Lenke: plink

Gender and/or age based author profiling

Until a few years ago, gender-based language studies mainly concentrated on speech. However, social media texts now provide plenty of data for extracting author profiles based on parameters such as gender, age and geolocation, while on the other hand posing new challenges for language analysis due to the often unconventional and abbreviated language used, as well as other characteristica of social media text such as usage of hashtags, emojis, emoticons, code-switching (mixing languages), etc. In addition, many users do not volunteer their actual and true profiles. The theme for this thesis project would thus be investigate automatic (machine learning) methods to either extract and classify author profiles in online texts, or to figure out whether one specific user (or type of user) could have written a given text (i.e., cyber forensics).

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Gender bias in texts and text analysis

Gender and other demographic biases in machine learned models have received increased attention in recent years. Language models and language analysis have been shown to be highly affected by such biases, which has triggered research on fair representations of gender in language models, aiming both at building and using fairer training and evaluation datasets, and at changing the actual learning algorithms themselves. There is still a lot of room for improvements though, both in those methods and in ways to quantify bias as such.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Identifying characteristics of persons vulnerable to social media extremism

Extremist groups take to social media since they facilitate cheap, quick, and broad dissemination of messages, and allow for unfettered communication with an audience without the filter or ‘selectivity’ of mainstream news outlets. There have in recent years been substantial efforts to identify members already belonging to extremist organisations and track their Internet activities. The present proposal, however, is primarily aimed at the individuals targeted by the extremists, i.e., persons susceptible to their ideas. The goals would be to profile persons vulnerable to extremism and intercept them before they fully turn to the extremist organisations, and to identify sources of extremism and hate-speech in order to preventively destabilize extremist networks.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Identifying Online Hate Speech and Cyber Bullying

During the Spring of 2017, parliamentary committees in Germany and the UK strongly criticised leading social media sites such as Facebook, Twitter and Youtube for failing to take sufficient and quick enough action against hate-speech, with the German government threatening to fine the social networks up to 50 million euros if they continue to fail to remove hateful postings within a week.
With legislation in other countries set to follow, properly identifying hatespeech is a pressing issue, not only for the major players, but also for smaller companies, clubs, and organisations that allow for user-generated content on their sites. Many such sites currently use slow, manual moderation, which mean that abusive posts will be left online for too long without appropriate action being taken or that content will be published with delay (which might be unacceptable to the users, e.g., in online chat rooms).

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Kahoot! Content understanding and answer suggestions

High-level description
This thesis will be carried out together with Kahoot! With over 5 billion cumulative players in the Kahoot! ecosystem, the amount of content stored on the platform is enormous. Kahoot! wants to enrich the content on the platform to build a better foundation for functionality like recommendations that will help improve end users’ experience.

Status: Valgbart     Egnet for: En student     Lenke: plink

Music emotion recognition and composition

The thesis project would explore either music emotion recognition or automated music composition based on emotion-annotated music - or a combination of those two themes. In either case the work needs to explore (general) emotion taxonomies, music data with and without emotional annotations, and techniques for music emotion classification.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Native Language Identification

Native Language Identification is the task of identifying the native language of a writer based solely on a sample of their writing in another language. The task is typically framed as a classification problem where the set of native languages is known beforehand. Most work has focused on identifying the native language of writers learning English as a second language. The master thesis work connects to previous work in IDI's AI group and potentially involves participation in a "shared task competition" on Native Language Identification where training and test data is made available by the organisers (such as https://sites.google.com/site/nlisharedtask2013/).

Status: Valgbart     Egnet for: En student     Lenke: plink

Plagiarism Detection

The task of plagiarism detection is to look at a given a document and establish whether it is an original or not. It can be further divided into two sub-tasks: source retrieval and text alignment. The aim of the thesis work would be to solve one or both of them (tentatively depending on whether one or two students worked together). In source retrieval, the task is: given a suspicious document and a web search API, retrieve all plagiarized sources while minimizing retrieval costs. While the task of text alignment is: given a pair of documents, identify all contiguous maximal-length passages of reused text between them.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Sentiment Analysis in Tweets

In recent years, micro-blogging has become prevalent, and the Twitter API allows users to collect a corpus from their micro-blogosphere. The posts, named tweets, are limited to 140 characters, and are often used to express positive or negative emotions to a person or product. In this project, the goal is to use the Twitter corpus to do sentiment analysis, that is, to classify tweets as to whether they express positive or negative opinions, or are neutral/objective. The work could build on previous master theses at NTNU, and potentially aim to participate in a shared task competition on Twitter Sentiment Analysis.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Sentiment Analysis of Figurative Language in Twitter

The master thesis project is aimed at the automatic classification of tweets containing figurative language, that is, language which intentionally conveys secondary or extended meanings (such as sarcasm, irony and metaphor). Such figurative language creates a significant challenge for sentiment analysis systems, as direct approaches based on words and their lexical semantics often are inadequate in the face of indirect meanings. One goal of the project is to find a set of tweets that are rich in figurative language, another goal is to determine whether the writer of each such tweet has expressed a positive or negative sentiment, and possibly the degree to which this sentiment has been communicated.

Status: Valgbart     Egnet for: En student     Lenke: plink

Societal Sentiment Analysis: Predicting Social Media Personalities, Values and Ethics

Several models can be used to find out how users’ social media networks, behaviour and language are related to their ethical practices and personalities, Such models include Schwartz’ values and ethics model and Goldberg's Big 5 model that defines personality traits such as openness, conscientiousness, extraversion, agreeableness and neuroticism. The thesis project would investigate applying such models to social media text and how the user personalties are reflected by the social networks that they participate in and develop.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Text-to-image synthesis

Text-to-image synthesis is a multi-modal task aimed at bridging the gap between text and image representations. State-of-the-art systems are reasonably successful at generating images of single objects, but struggle to generate images of more complex scenes (e.g., of interacting objects) and higher resolutions images. Currently utilised techniques are mostly based on Generative Adversarial Networks (GANs), where one neural network generates images while a second network (the descriminator) tries to separate generated images from genuine ones, with both networks aiming to improve over time based on the their success. Within text-to-image generation, the generator can get additional feedback through re-generating of text from the generated images. The thesis project would target further improvements in the field, in particular the generation of complex scenes at high resolutions, by investigating better text analysis and re-generation techniques.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Using Evolutionary Algorithms to Investigate Language Evolution

The aim of this project will be to use evolutionary algorithms as a vehicle to investigate the main dynamics in language evolution. Language evolution is a highly multi-disciplinary research field with the main theories involving biological evolution, language learning, and cultural evolution. When trying to understand the origins of languages, we can try to compensate for the lack of empirical evidence by utilising evolutionary computational methods to create simulations of how language may have evolved over time, e.g., by creating "language games" to simulate communication between agents in a social setting. In general, simulations on language evolution tend to have relatively small and fixed population sizes, something this study could aim to change.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

A Tool for In-Video Feedback: Design and Evaluation

Supervisors: Kshitij Sharma, Michail Giannakos
Place: LCI Lab: https://lci.idi.ntnu.no/

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Adaptive and Gamified Learning Technologies to Support Motivation and Engagement

Supervisors: Zacharoula Papamitsiou, Michail Giannakos, Alf Inge Wang
Place: LCI Lab: https://lci.idi.ntnu.no/

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Adaptive Teaching Technologies to Create and Monitor Learning Activities

Supervisors: Zacharoula Papamitsiou and Michail Giannakos
Place: LCI Lab: https://lci.idi.ntnu.no/

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

AI and Multimodal Data to Monitor and Support Children’s Learning Activities

Supervisors: Michail Giannakos, Kshitij Sharma and Serena Lee-Cultura

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Collaborative code editing tool: Design and Evaluation

Supervisors: Kshitij Sharma, Michail Giannakos
Place: LCI Lab: https://lci.idi.ntnu.no/

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Generalizability of eye-tracking and EEG features

Supervisors: Kshitij Sharma, Michail Giannakos
Place: LCI Lab: https://lci.idi.ntnu.no/

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Multimodal Learning Analytics for video-based learning

Supervisors: Kshitij Sharma, Michail Giannakos
Place: LCI Lab: https://lci.idi.ntnu.no/

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

The design and use of AI-enabled learning systems

Supervisors: Ilias Pappas & Michail Giannakos

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Utilizing Electromyography (EMG) as an Input Modality for Head-Mounted Displays (HMDs)

Supervisors: Evangelos Niforatos and Michail Giannakos

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Wearable Memory Augmentation: Moving Cued Recall out of the Lab

Supervisors: Evangelos Niforatos and Michail Giannakos

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

[NorwAI] Value of AI Technologies

AI and Big Data technologies are central in many digitalization or digital transformation projects. They are sometimes used to cut costs of existing services, but the technology is often adopted to offer improved or brand new services. There are today best practices from many industry sectors, and practitioners are researchers have worked out guidelines for the integration of AI in organizations. However, many projects suffer from a lack of understanding how the nature of AI affects technical infrastructures and business strategies, and social and ethical implications are often underestimated and not sufficiently addressed.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NorwAI/Cognite] Explaining Deep Learning Models for Entity Matching

In the master project the student(s) will explore different techniques for explaining machine learning models and how they can be applied to entity matching. After initial exploration we expect the student(s) to focus on one particularly promising technique and perform extensive quantitative and qualitative evaluation on established public entity matching datasets to discover strengths and weaknesses. Finally, in a possible follow-up master thesis the student(s) will suggest and evaluate ways to improve the selected technique.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NorwAI/DNB] Conversational robots

Conversational agents (chatbots) are systems that can enter into dialogues with real users.  Most current chatbots are rule-based, but some newer systems are based on generative models and machine learning. 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NorwAI/Sparebank1] Aggregation of input variables in machine learning

This project is about finding the optimal aggregation of input variables in machine learning systems at Sparebank1 SMN.  

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NorwAI/Sparebank1] CRM Text Analysis

Customer interaction with Sparebank1 SMN is recorded as user-written free text. The project will develop AI/ML applications to structure and extract information from these texts for the following purposes:

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Active learning for industrial entity matching with machine learning

Cognite are bringing modern cloud software and data analysis to many industrial companies. A common obstacle for industrial companies is that they have data residing in many different systems without common identificators (equipment, time series, 3D models, documents, etc). To gain value from digitalization it's important to be able to work across several data sources - e.g. click on something in a 3D and see its sensor time series. Entity matching is the problem of deciding which records from two different data sources refer to the same real-world entity.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Conversational Approach to News

Traditional media houses publish news stories that are either updated or replaced with new stories as events are unfolding. The stories are presented as complete texts that are supposed to be read from beginning to end. For small devices like mobile phones it may be interesting to look into other ways of presenting news stories. In some recent experiments news stories have been broken up into several pieces that have either been structured as a conversation or presented piece by piece by avatars.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Discovery and scoping of data for industrial entity matching

Cognite is bringing modern cloud software and data analysis to many industrial companies. A common obstacle for industrial companies is that they have data residing in many different systems without common identificators (equipment, time series, 3D models, documents, etc). To gain value from digitalization it's important to be able to work across several data sources - e.g. click on something in a 3D and see its sensor time series. Entity matching is the problem of deciding which records from two different data sources refer to the same real-world entity.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Explainable Recommender Systems

In news recommendation both collaborative filtering and content-based recommendation techniques are used. Often contextual features like time and location, as well as additional data from social media, are also taken into account to personalize news services to each individual reader. Due to the complexity and number of recommendation strategies employed, it is difficult for the reader to understand the logic behind the recommended articles. The lack of transparency hampers readers’ trust in the system and makes it hard to detect flaws in the recommendation engine.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Explaining Fake News

Fake news stories are stories that deliberately try to disinform readers or spread hoaxes via traditional media channels. Their intent is to mislead in order to damage parties, people or organizations, or for financial gains. Recent work on network analysis, linguistic properties and classification has tried to use machine learning and other techniques to detect fake news stories by contrasting them with verified true stories. Unfortunately, even if a fake story is identified, the techniques have not been very good at explaining why the story is considered fake.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Industrial entity matching on documents with machine learning

Cognite are bringing modern cloud software and data analysis to many industrial companies. A common obstacle for industrial companies is that they have data residing in many different systems without common identificators (assets, time series, 3D models, documents, etc). To gain value from digitalization it's important to be able to work across several data sources - e.g. click on something in a 3D and see its sensor time series. Entity matching is the problem of deciding which records from two different data sources refer to the same real-world entity.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Industry friendly modeling for labeled property graphs

Cognite is delivering digital twin and industrial IOT software to industrial companies across manufacturing, oil&gas and power&utilities. The core of Cognite's offering is called Cognite Data Fusion(CDF) and is powered by a labeled property graph concept that enables flexible modeling of industrial information systems for small and larger companies to collaborate in projects and in operations.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Interactive industrial entity matching with machine learning

Cognite are bringing modern cloud software and data analysis to many industrial companies. A common obstacle for industrial companies is that they have data residing in many different systems without common identificators (assets, time series, 3D models, documents, etc). To gain value from digitalization it's important to be able to work across several data sources - e.g. click on something in a 3D and see its sensor time series. Entity matching is the problem of deciding which records from two different data sources refer to the same real-world entity.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Reinforcement learning: Personalizing a virtual driving instructor

Problem description
Way AS research and develops a virtual driving instructor that provide instructions to students learning to drive in a full-scale car simulator. Learning should be adjusted to each individual student.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Robust time-series forecasting

Problem description

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

TrønderEnergi: Explainable time-series forecasting

Problem description
TrønderEnergi has several forecasting systems utilizing machine learning methods in commercial operation. The systems are used and the forecasting is monitored by domain experts in the operating center at Berkåk. The domain experts do not have any training and knowledge in the machine learning methods, but they do have to make decision based on their output. At times the experts need to understand why a forecast has the value it has. The literature has to be reviewed, prototypes should be developed and evaluated by domain experts.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

TrønderEnergi: Reproducibility of deep transfer learning models for time-series forecasting

Problem description
TrønderEnergi is conducting research on both deep transfer learning models for time-series forecasting and reproducibility. Reproducibility is a challenge for many machine learning models. Deep learning models have specific challenges in this regard. No studies have been done on the reproducibility of transfer learning models for time-series data.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Event Mining and Linking

News publishing has seen exponential growth over the last decade as any user
on the Web can create newsworthy information. To understand such minute and
specialized information (e.g., "Ian Nepomniachtchi qualified for the World
Championships") often contextualization with prior related information is
required (e.g., "Magnus Carlsen is the undefeated chess world champion since
2013", "World Championship matches were halted due to Covid-19 in 2020", and
"Nepo and Giri battle for the contendership to challenge Carlsen in the now
resumed World Championship in Russia.") In this project we will look at ways
we can mine events from large news archives and weave them together into
storylines for end-user comprehension and consumption.

Broad research areas: data mining, databases, information retrieval, and
natural language processing.

Number of students: 2.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Learning Indexing Patterns from Annotated Document Collections

Large document collections can now be annotated with entities, temporal, and
numerical expressions quite readily with the help of natural language
processing tools (e.g., Stanford CoreNLP toolkit). It has been shown that
words conveying the semantics of such annotations form a large chunk of query
logs. For example, "tokyo olympics" refers only to the Olympics taking place
in Tokyo whereas "olympics 1990s" refers to the Olympics taking place during a
particular time period. Similarly, "us presidential elections 2000s" refers to
specific entities and particular time periods of interest. In this project, we
will take a closer look at how we can index the annotated document collections
in such a way that selecting relevant documents from certain time periods,
entity categories, and numerical ranges becomes easier.


Broad research areas: databases, information retrieval, data mining, and
natural language processing.

Number of students: 1.
 

Status: Valgbart     Egnet for: En student     Lenke: plink

Learning Indexing Patterns from Dependency Parse Trees

We can annotate sentences in documents with dependency parse structures that
make explicit the linguistic roles of the words within a sentence as well as
relationships amongst themselves. Such parse structures can help us extract
relevant fragments of the sentence connected to an entity. For instance,
extract all relevant relationships associated with the entity "Norway" to
other entities (persons, organizations, or locations) (e.g., "Norway's
government offices are located in Oslo"). Such information extraction
techniques are quite useful in building of knowledge graphs and are often done
by specifying extraction patterns over document collections manually. In this
we will look at how we can do this at scale by learning extraction patterns
and indexing of dependency parse trees.

Broad research areas: databases, information retrieval, natural language
processing, and machine learning.

Number of students: 2.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Numerical Contours

In this project, we will looking at how we can construct a scalable annotator
for temporal and numerical expressions in large document collections. For
instance, consider Wikipedia as a concrete collection of documents. As an
example consider annotating the Wikipedia entry of "Tesla_Inc." with the
following excerpt:

Status: Valgbart     Egnet for: En student     Lenke: plink

Semantic Text Indexing

Popular deep learning techniques such word2vec allow words to be represented
as vectors that allow semantic text matching. As a concrete example, when
searching for documents containg the keywords "installing linux on machines
with graphic cards". For such a query we will not get many relevant documents
as many categories of the query are underspecified (e.g., "linux", "machines",
and "graphic cards"). This problem at a rudimentary level can be solved with
pseudo-relevance feedback. But can we do better? One approach is using deep
learning methods (e.g., word2vec). However, as with many deep learning
methods such methods are often treated as blackboxes and leave little room for
implementing such methods at scale. In this work, we will look at alternative
ways of endowing text with semantics and perform semantic text matching at
scale across millions of documents.

Broad research areas: databases, information retrieval, and natural language
processing.

Number of students: 1.

Status: Valgbart     Egnet for: En student     Lenke: plink

Text Driven Analytics and Visualization

The Web serves a large repository of human knowledge in the form of text
documents over a substantial period of time. We can now obtain reliable
annotations in the form of entities, time and numbers over millions of
documents. In this project, we will design a system that leverages a scalable
information extraction infrastructure for generating interesting
visualizations and allow the user to perform text-driven analytics.

Broad research areas: databases, information retrieval, and data mining.

Number of students: 2.
 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Bio-inspired techniques for remote sensing

Our planet is changing, now more than ever before. Understanding these changes and how they impact the environment is crucial for preserving the Earth for the coming generations.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

bio-inspirerte metoder

Oppgavene skreddersys til studenters interest.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Development of a novel immune-inspired hybrid classification algorithm

The immune system is arguably one of nature's most highly adaptive, distributed and self-organising systems. It has the property of being able to recognise anomalies --- something that deviates from the common rule.  In 2018-2019 two masters students proposed a novel hybrid classification algorithm MAIM, combining such features of the Immune Systems with an Island Model Genetic Algorithm (IGA). The preliminary results achieved are promising and resultet in an international publication. The goal of this project is to build on this foundational work  and investigate the many avenues available to further extend/refine this novel algorithm. 

Contact Pauline haddow, pauline@ntnu.no for further information

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Particle Swarm Optimisation as a modelling tool for Climate Change

Particle Swarm Optimisation, where each particle may be thought of as an individual in a society of particles, provides a  technique to model influences and effects in society that may prove beneficial  in the battle against Climate change.  This project is focussed on developing a PSO modelling environment that enables the study how an individual may become more inclined to increase their individual contribution/sacrifice for the benefit of future populations. 

contact: Pauline Haddow

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

VRP with trucks and drones for Health Service package delivery

In the last years, unmanned aerial vehicles (drones) have attracted the interest of both academia and industry due to their potential to change the way transportation and logistics are tackled. Drones have the potential to significantly reduce the cost, time and reliability of last-mile deliveries. In order to manage transportation by road or air or a combination of both, a vehicle routing problem (VRP) needs to be solved. The vehicle routing problem with drones (VRPD) is an extension of VRP, where drones or a tandem strategy of trucks and drones are involved in the delivery of parcels to customers. One application area of interest is the delivery of small medical packages to inaccessible, remote or dense urban areas where such packets may include blood samples, medicines and vaccines. Such VRPD need to take into account the dynamic and uncertain nature of the application area.
Various masters projects are available within this topic and may include the application of various biological-inspired algorithms such as Genetic algorithsm, Multi-objective Optimisation or Particle Swarm Optimisation.
Contact: Pauline Haddow, pauline@ntnu.no

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Algorithmic problems

Several problems related to algorithms and data structures may be used as the basis for projects and masters theses.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Algorithmic problems

Several problems related to algorithms and data structures may be used as the basis for projects and masters theses.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Dronebasert sanking av sau

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Gjenfinning av sau ved hjelp av drone

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Manuell oppfølging av sau på beite

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Sau og rovdyr

Sau slippes ut på beite på våren og går på beite i ca 16 uker. Bonden skal holde ukentlig oppsyn med de i denne perioden. Mye sau har i dag en radioenhet som daglig rapporterer hvor de er. Rovdyr kan angripe, skade og drepe sau på beite. Vi er her interessert i å detektere om det er rovdyr i sauens beiteområde basert på analyse av sauens bevegelsesmønster. Bevegelsesmønsteret i et område hvor det forekommer rovdyr vil avvike fra sauens normale bevegelsesmønster. Dette avviket er vi interessert i å detektere så tidlig som mulig slik at sauebonden har mulighet til å sette inn tiltak for å få sauene ut av området før tap oppstår. Vi har fått tilgang til bevegelsesdata gjennom hele beitesesongen for sau fra et antall bønder. Disse data vil benyttes i analysen.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Sporing av sau ved hjelp av enkel radioteknologi

Masteroppgave:
Sporing av sau ved hjelp av enkel radioteknologi

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Analysing students' feedback with ENA to improve teaching and learning

Epistemic Network Analysis (ENA) can be a useful tool in analysing connection strength between coded data such as aspects and topics identified from users’ feedback for a given target subject. This research project aims at exploiting ENAs capability in answering questions related to the relationship between various aspects and the connection strength between them from student’s feedback on a course. This work is based on an existing MOOC dataset for the sentiment analysis task carried out by master students at NTNU.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Assessing students’ attitudes across various learning disciplines using sentiment analysis

Sentiment analysis is an effective tool which has been widely used in education domain to examine students’ reviews to improve learning and teaching. Students can have different attitudes and sentiments towards various learning disciplines. For example, ones may show more interest (positive attitude) towards ‘Engineering’ and less interest (negative) towards ‘Social sciences’. Therefore, the aim of this research project is to assess students’ satisfaction and attitudes across different disciplines using aspect-based sentiment analysis. Data needed to conduct the study should be collected from MOOCs of different disciplines such as social sciences, natural sciences, engineering, etc.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Big data analytics in medical imaging using deep learning

Artificial intelligence (AI) continues to garner substantial interest in medical imaging. The potential applications are vast and include the entirety of the medical imaging life cycle from image creation to diagnosis to outcome prediction. In addition, big data has flattering more dominant in healthcare, due to three major reasons, such as the huge amount of data available, expanding healthcare costs, and a target on personalized care. Big data processing in healthcare refers to generating, collecting, analyzing, and holding clinical data that is too vast or complex to be inferred by traditional methods of data processing methods. Big data sources for healthcare include, the Internet of Things (IoT), Electronic Medical Record/Electronic Health Record (EMR/EHR) contains patient medical history, diagnoses, medications, treatment plans, allergies, laboratory and test results, genomic sequencing, Medical Imaging, Insurance Providers and other clinical data.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Deep Active learning assisted Wild Animal Classification methods for camera trap in large Scale

In recent years the world’s biodiversity is declining on an unprecedented scale. Many species are endangered and remaining populations need to be protected. To overcome this agitating issue, biologist started to use remote camera devices for wildlife monitoring and estimation of remaining population sizes. Unfortunately, the huge amount of data makes the necessary manual analysis extremely tedious and highly cost intensive. This presents a major obstacle to scientists and ecologists to monitor wildlife in an open environment. Therefore, in this research we offer to propose a novel approach to organize the most challenging Snapshot Serengeti dataset and then classify the wild animal species using Deep Active learning methods.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Deep learning enabled IoT Devices for Intelligent decisions in Multimedia Surveillance Networks

Wireless multimedia surveillance networks produce large amounts of visual data. Lack of intelligent mechanisms for storage, indexing, and management makes these tasks comparatively, time consuming for analysts to browse and gather actionable intelligence. Taking advantage of the powerful feature of fog computing over economical hardware (i.e., Jetson Nano), the purpose of this work is to propose a mechanism to process surveillance data (i.e., multimedia contents and contextual information under uncertain environments) at visual processing hub and only transmit relevant and useful data to Base-Station securely. In result, the proposed method minimizes energy and bandwidth consumption and ensure effective surveillance while extending its potential toward the ongoing projects related to surveillance in Norway.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Human Action Recognition using Deep Learning Assisted Visual Attention Model

Human activity recognition (HAR) aims to recognize activities from a series of observations on the actions of subjects and the environmental conditions. HAR is an important challenge in a variety of applications including health care, human-computer interaction and intelligent video surveillance to enhance security in different domains. The evaluation of the proposed algorithm depends on: 1) training data, and 2) learning model. The aim of this work is to investigate the effectiveness of deep learning models meeting the challenges posed by existing state-of-the-art techniques over different publicly available datasets while replicating human visual attention model.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Image Generation and Clinical Assessment

In todays health care, imaging plays an important role throughout the entire clinical process from diagnostics and treatment planning to surgical procedures and follow up studies. Since most imaging modalities have gone directly digital, with continually increasing resolution, at the same time, medical image analytics have many challenges. Firstly, there is a need for more generic image analysis technologies that can be efficiently adapted for a specific clinical task. Secondly, efficient approaches for ground truth generation are needed to match the increasing demands regarding validation and machine learning. Thirdly, algorithms for analyzing heterogeneous image data and their fusion are needed . This research work aims to contribute to ongoing research in terms of more accurate and robust algorithmic solutions

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Intelligent Camera Prioritization using Violent Activity Recognition in Large Scale Surveillance Networks

Digital surveillance systems are universally installed that continuously generate a massive amount of data, which require manual monitoring to recognize uncertain human activities in public areas. Intelligent surveillance systems are highly desirable that can automatically identify normal and abnormal activities to allow efficient monitoring by selecting only those cameras, where abnormal activities occur. The purpose of this work is to proposes an energy-efficient camera prioritization framework that intelligently shares priority between cameras in a vast surveillance network using feedback of the activity recognition system. The proposed work consists of: 1) addresses the limitations of existing manual monitoring surveillance systems, 2) select the salient frames from the online video stream, 3) compute a feature map using lightweight deep learning model to rank the priorities of the cameras, 4) the probabilities predicted in step 3 and metadata of the cameras are processed to compute the priorities, 5) selection of the camera based on step 4

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Multi label classification report of medical image (X-rays, MRI and CT) Translation

Huge data in the form of raw text is available in radiologists’/specialists’ clinical notes, where its rough nature makes it challenging for meaningful information extraction such as specific disease names, entity recognition, and concept extraction. The amount of data increases daily, thereby requiring efficient browsing and searching techniques for effective and useful contents extraction that is utilized in further research. Clinical Natural language processing (CNLP) is a prominent solution in this direction, dealing with concept extraction from clinical text, diseases identification and early predictions of medical outcomes based on medical history. A handsome amount of techniques is developed using CNLP strategies, but still there is a space for effective and efficient methods that are precise and trustworthy for real-world implementation.

Status: Valgbart     Egnet for: En student     Lenke: plink

Pedagogical chat bot

The task involves developing a domain specific conversational AI bot for a Hyper Interactive Intelligent Presenter. The bot would be trained to converse on a domain-specific topic for a course and will be useful to assist user in acquiring basic knowledge about a topic and/or for querying relevant material available on the course page. The bot is to be trained using deep learning on a variety of topics within a domain and be evaluated against different queries.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Synthetic sample for extremely imbalance data

Data imbalance is a frequently occurring problem in a classification task where the number of samples in one class exceeds the amount in other classes. Quite often, the minority class data is of great importance representing concepts of interest and is diffiuclt to obtain in the real dataset. Lack of enough data samples results in data imbalance causing poor classification performance while training. Synthetic data generation techniques such as SMOT can address this issue, yet such methods suffer from overfitting and substantial noise. This research work aims at creating an efficient data generation technique overcoming challenges posed by existing state-of-the-art methods.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

AI applications for social good (AI4SG)


Supervisors: Ilias Pappas, Letizia Jaccheri

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Developing Software for inclusion with new technological trends such as AI and IOT

The purpose of this thesis, that can be chosen as project and as thesis, by one or two students and also by several groups
is to develop a new Software product that incorporates novel technology and at the same time targets stakeholders that
traditionally have not been included as the primary software stakeholders. This can be children, or women, or people
with disabilities.
Software engineering is the research field which focuses on establishing new knowledge on key factors for the successful
development of some of the most complex systems ever developed by humans [1].
The current software engineering body of knowledge is challenged by new technological and societal development.
Technological developments like Artificial Intelligence, Cloud storage and processing, the Internet of Things, cyberPhysical systems and big data are changing the nature of how software is conceived, developed, and used. At the same,
software pervades all aspects of society. In today’s digital, automated and globally connected society, software is used
by larger and larger part of the world population.
Massive and ever growing amount of software is available today to different people of younger and younger age through
sites such as Facebook and Instagram, apps, and games. People are increasingly using software intensive technologies.
In some cases, the software is specifically made for solving a problem, like helping people with reading difficulties to
learn how to read [2] or people with obesity to exercise [3]. In most cases, software is made for commercial purposes.
It is getting difficult every day to develop these tools, apps or websites, especially for people with special needs.
In previous studies we have addressed creativity software for children [4]; exergames [3]; and software for motivating
adolescents with Intellectual Disabilities [5] [6] and for increasing empathy[7].
An emphasis will be put on studying existing software for inclusion and develop a set of enablers (or characteristics) for
good software for inclusion. Ethics, privacy, explanability, cybersecurity, and user well-being will have to be integral
part of the processes, so that the people who use software systems will be offered new possibilities.
The student (or couple of students) will study a sub theme of the theme SE for inclusion, study the literature, plan the
development of the software and its evaluation.
The student(s) will:
• Define a research question.
• Run a literature review and identify similar software applications 
• Plan, development and testing of a software application in several iteration - see [8][9]
• Document the development and data analysis to propose lesson learnt about software development for inclusion
• Paper Writing for Conference (if the student is interested in publishing)

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Investigating Gender and Diversity in Software Development

This project/master thesis will build on existing bulk of knowledge about gender and diversity in software development (TDT10) to provide increased knowledge and solutions based on empirical studies with Norwegian and International  IT industry, or public sector, or entrepreneurial companies.
Specifically in this project/master thesis, the student(s) will propose one or more goals to investigate.
The general research question is “What is the relation between Software Engineering Processes and SDG5”?

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Investigating Gender and Diversity in Software Development - 2

This project/master thesis will build on existing bulk of knowledge about gender and diversity in software development (TDT10) to provide increased knowledge and solutions based on empirical studies with Norwegian and International  IT industry, or public sector, or entrepreneurial companies.
Specifically in this project/master thesis, the student(s) will propose one or more goals to investigate.
The general research question is “What is the relation between Software Engineering Processes and SDG5”?

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Software Engineering and Smart Society

The purpose of this thesis, that can be chosen as project and as thesis, by one or two students and also by several groups is to support the implementation of AI software inside the Software Development process. Each student (or couple of students) will study a sub theme of the theme SE and AI, study the literature, plan an empirical investigation to collect data from processes that involve software engineers and software companies.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Software for Children and Teens

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Software for Children and Teens 2021-2022

The purpose of this thesis, that can be chosen as project and as thesis, by one or two students and also by several groups is to increase knowledge about Software Engineering for and which children. Each student (or couple of students) will study a sub theme of the theme SE and children, study the literature, plan an empirical investigation to collect data from processes that involve children, software and possibly software engineers.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Software4Good: Understanding the relation between Sustainable Development Goals and Software Engineering Processes

Nowadays, the world faces pressing social and environmental issues and both Science and software has an important role to play in SDG achievement.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

The role of employee-driven innovation within private and/or public organizations


Supervisors: Leif Erik Opland, Ilias Pappas, Letizia Jaccheri

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Understanding Software for Children and Teens

The purpose of this thesis, that can be chosen as project and as thesis, by one or two students and also by several groups is to increase knowledge about Software Engineering for and which children. Each student (or couple of students) will study a sub theme of the theme SE and children, study the literature, plan an empirical investigation to collect data from processes that involve children, software and possibly software engineers.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Designing an intermittent computing demonstrator

In an intermittent computing system, computation occurs in response to energy becoming available. These systems are typically used in scenarios where changing batteries is either impossible or expensive -- or environmental constraints means that chemical batteries cannot be used. Intermittent computing systems hence must support wireless communication, which is inherently challenging without a stable energy supply.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Domain Specific Accelerators (DSAs) in FireSim

FireSim (see fires.im/) is the state-of-the-art FPGA-accelerated computer architecture simulator. Since it is capable of simulating complete benchmarks, it is particularily well suited to study the overheads of invoking Domain Specific Accelerators (DSAs). The project will hence be about identifying suitable DSAs (preferably ones that already exist), understanding out to best integrate them (both in terms of software and hardware).

If this sounds cool, let me know and I'll provide more details. 

 

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Evaluating Quantum Computing Simulators

Recent advances in the construction of quantum computers indicate that practical quantum computers may soon become a reality. In the meantime, we need to use simulators to perform architectural exploration of quantum computing design options. A plethora of simulators exist [1], and the student is expected to survey the available simulators to understand which simulators are most suitable for quantum computer architecture research. After identifying one or more simulators, the student should assess their architectural fidelity as well as their scalability. This will then (hopefully) create the foundation for a master thesis where we attempt to improve the scalability of quantum computer simulators.

[1] quantiki.org/wiki/list-qc-simulators

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Towards simulating GPU compute workloads in FireSim

FireSim [1] is an FPGA-accelerated computer architecture simulator infrastucture. Unfortunately, it currently does not support simulating CPU+GPU heterogeneous systems. The objective of this master thesis is to investigate how such systems could be simulated and (ultimately) bring up a proof of concept implementation -- which will require lots of interesting RTL-level implemenation using Chisel.

Sounds interesting? Let me know, and I'll provide more details.

[1] https://fires.im/

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

AI automated cardiovascular therapy and diagnostics

For cardiovascular therapy as well as valve repair or replacement, a hybrid operating room is becoming a prerequisite. To facilitate a correct device deployment the relationship between the anatomy and the position of the instruments has to be known and presented to the surgeon. The position of the instruments and the geometry of the vascular tree is best visualized on Xray images, with a wide field of view, however the resulting images are 2D projections of the 3D world for a given a position of the X-ray source. On the other hand the heart’s anatomy is very well depicted using ultrasound images (e.g. 3D transesophageal ultrasound - TEE), however these images have a much smaller field of view due to the positioning of the TEE probe with regards to the heart. Therefore, tools that can automatically identify the position and orientation of the ultrasound probe and cathethers in X-ray images as well as automatic localization of anatomic structures of interest is highly desirable. Today most of these tasks are solved manually or semi-automatically, however automating them would improve the accuracy and also reduce the amount of time the interventionalists spend on aligning the images.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Augmented reality tools for cardiac and bronchoscopic procedures

The aim of this work is to develop a new visualization framework using mixed reality which may improve the way data is presented to an operator during a cardiac ultrasound exam or bronchoscopy procedure.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Remote Learning and Collaboration with Augmented Reality and Hololens 2

Augmented Reality (AR) can provide rich and interactive learning experiences and performance augmentation for remote distributed learners. The use of AR for collaboration and learning has significantly increased during the ongoing pandemic and has been widely adopted by several companies (e.g. Equinor), hospitals treating COVID-19 patients (https://www.businessinsider.com/london-doctors-microsoft-hololens-headsets-covid-19-patients-ppe-2020-5?r=US&IR=T) and educational institutions.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Towards a 3D model of the brain in AR as an educational tool

The aim of the MSc project is to investigate the feasibility to visualize the current version of the Waxholm 3D atlas in AR and to explore the applicability as an educational tool in settings where one teacher works with groups of students and students can subsequently explore brain anatomy independently.

 

Status: Tildelt     Egnet for: En student     Lenke: plink

Volume rendering on a mixed reality device

The project will investigate the possibility of implementing volume rendering techniques directly on a mixed reality headset (e.g. Microsoft HoloLens). Main purpose is to visualize patient-specific ultrasound data and/or computed tomography scans and combine them with geometric models representing structures of interest.

 

Status: Tildelt     Egnet for: En student     Lenke: plink

Kontinuerlig innhenting av brukertilbakemelding

I mange digitaliseringsprosjekter er det å kunne tilpasse funksjonaliteten i systemene til brukerkrav og brukskonteksten en utfordring. For å skape en bedre forståelse av krav og kontekst i forkant av utviklingsaktivitetene benyttes det ofte brukersentrert design-metodikk i disse utviklings-prosjektene. Dessverre er det ikke alltid slik at problemer knyttet til funksjonalitet blir avdekket før systemet blir satt i drift [1].

Status: Valgbart     Egnet for: En student     Lenke: plink

Cool AI stuff

Bring your own idea. We bring AI.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

App for tidlig deteksjon av dysleksi

Forskning ved psykologisk institutt indikerer at en underliggende årsak for blant annet dysleksi og dyskalkuli kan være dysfunksjon i visuell prosessering. Nærmere bestemt er det nerveceller som flytter informasjon saktere hos disse gruppene enn hos de som ikke har disse utfordringene. Disse nervecellene kalles magnoceller og har i hovedsak ansvar for å oppfatte raske forandringer i omgivelsene.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Development and maintenance of IT-systems in Norwegian Organization

In 1993, 1998, 2003, 2008, 2013 and 2018  surveys were performed to investigate development and maintenance of IT-systems in Norwegian organizations. Comparable data for important areas where also capture in collaboration with Rambøll IT I praksis investigations. Data collection for a similar survey is carried out yearly since  2015. The assignment will be to analyze the quantitative and qualitative data from recent investigations. Together with a literature review, the survey investigations are expected to give us new knowledge about mechanisms affecting resource utilization related to information systems support in organizations. The report should be written in English and is expected to form the basis for scientific publications

Status: Valgbart     Egnet for: En student     Lenke: plink

Digital registrering av innemiljø og tilstedværelse på Campus

I sammenheng med Campus-prosjektet instrumenteres ulike rom med sensorerer for inneklima og tilstedeværelse. Oppgaven går ut på å utforske data fra denne type sensorer i forhold til å finne gode anvendelser av disse.

Prosjektet vil også relatere seg til forskningsrådsprosjektet 'Smart workplaces past Covid-19 som er et samarbeid med Mazemap og CISCO

Status: Valgbart     Egnet for: En student     Lenke: plink

Disruptive technologies for energy trading platforms

In connection to the digital economy NTNU project (https://www.ntnu.edu/digital-transformation/digeco ) , we investigate platforms to support energy trading, and we have a project task in connection to this.

Status: Valgbart     Egnet for: En student     Lenke: plink

Quality of models and modeling languages

In connection to organizational changes (including systems development) modeling is used in different ways. A general framework for assesing the quality of models and modeling languages is developed (SEQUAL). This task is related to specialize the framework to different domains of modeling. Possible specialization areas are enterprise modeling, process modeling, requirements specifications, ontologies, design models, data (data quality), methods and interactive models. The direction of specialization is to be discussed with the candidate and potential external user interests.

The report is expected to be written in English.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Analysing the Instruction Cache Behaviour of Cloud Applications

The popularity of cloud applications (such as search, video streaming, social media etc.) has shot up in the last decade due to availability of powerful mobile devices and ware-house scale datacenters. From computer architecture perspective, the distinctive characteristic of these applications is their gigantic instruction footprints that reach into multiple megabytes[1]. The instruction cache found in modern processors, in contrast, are only few tens of kilobytes.

Status: Valgbart     Egnet for: En student     Lenke: plink

Exploiting Memory-Level-Parallelism in Out-of-Order Cores

The long latency of memory accesses continue to be a critical performance bottleneck in modern out-of-order (OoO) cores. The techniques that exploit memory level parallelism (MLP) try to mitigate the bottleneck by hiding the latency of later requests, at least partially, under the shadow of earlier requests.

Status: Valgbart     Egnet for: En student     Lenke: plink

Improving Branch Prediction Accuracy for Cloud Applications

Contemporary cloud applications (such as search, video streaming, social media etc.) feature massive instruction footprints stemming from deep, layered software stacks. As a side effect of these massive footprints, the branch working set of these applications is also gigantic which complicates detecting and predicting these branches accurately at the fetch stage. The two key components required for branch prediction are: 1) Branch target buffer (BTB) that detects branches before the instructions are even fetched and 2) Branch predictor that predicts the direction of conditional branches (taken/not-taken).

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Power Efficient Instruction Scheduling in Out-of-Order Cores

The end of Dennard scaling has made all computing systems power constrained. It means that we can no longer burn more power to achieve higher performance; rather we have to increase the power-efficiency of a system to achieve higher performance within the same power budget.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Uninterrupted Instruction Supply in Servers

Contemporary server workloads, e.g. web search (Google), video streaming (Youtube), etc., feature massive instruction footprints stemming from deep, layered software stacks. The active instruction working set of the entire stack can easily reach into multiple megabytes, resulting in frequent front-end stalls due to instruction cache misses and pipeline flushes due to branch target buffer (BTB) misses. A recent study [1] at Google shows that these front-end stalls already account for up to 30% of execution time. In addition, the instruction sets are growing at 25% per year; therefore, the bottleneck is going to be even severe in future.

Status: Valgbart     Egnet for: En student     Lenke: plink

[EXAIGON]: Explainable AI (XAI) for power bids -- with TrønderEnergi

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[EXAIGON]: Explainable AI for preventive maintenance -- with TrønderEnergi

Introduction
TrønderEnergi owns a number of wind production parks, and also runs other parks as a subsidiary. Wind-power production facilities are complex systems, with rotating parts that sometimes break and therefore forces halts in production. To avoid unforeseen lost production, and to run the power production as efficiently as possible, TrønderEnergi make preventive maintenance plans based on data for typical life-length of equipment and rules of thumb. The AI department at TrønderEnergi is currently developing a “black box” monitoring system that will incorporate real-time measurements of working conditions (e.g., temperature, weather, etc). The goal is to avoid unplanned shut-downs through maintenance plans that fix or replace deteriorated parts at opportune times (e.g., when the production is low due to lack of wind).

Problem description
The black-box system for maintenance planning is central in TrønderEnergi’s operation of the wind-mill parks. It will be continuously confronted with a diverse set of end-users:

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NorwAI]: Explainable AI (XAI) -- with Sparebank 1 SMN

Sparebank 1 SMN is interested in developments in the field of explainable AI in general, and any innovations that can “open” the black box to some extent.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NorwAI]: Fair recommenders -- with Schibsted

Introduction
The E-Commerce and Distribution team is a business unit within Schibsted responsible for all companies that offer e-commerce and parcel delivery services. The two best known companies within the group are Helthjem and Morgenlevering, which have experienced great growth particularly after the pandemic.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Fantasy Premier League

Fantasy Premier League (FPL) is a popular game in which you set up a team of footballers from the English Premier League, choosing your players within a specific budget.  Each game-week you'll get feedback on how your players performed in terms of points. You are then given  the opportunity to trade players before next game-week starts (you have one "free" trade, that can be saved up for later, and also the opportunity of making additional trades, but at the expense of point-deductions). The players' values can change from one game-week to the next, hence there is a possibility to gain (or loose) funds to be invested in new players. You need to cover all playing positions in the squad and in the team, following a certain rule-set, and so on. Finally, there are lots of extras, like choosing one player as your captain, and using so-called wildcards in clever ways (see the website for detailed rules).

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Prosjekt for Mattis og Sebastian

Finans-relatert.

 

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Spilleroverganger

Prosjektet går ut på å bruke dataanalyse som beslutningsstøtte i arbeidet med å identifisere nye spilleroverganger for fotballklubber. 

Dette prosjektet er for Eliot Karlsen Strobel og Markus Malum Kim

 

 

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Uncertainty-aware Deep Learning a.k.a. Deep Bayesian Learning

Description:
Deep learning has been reported to improve upon previous state of the art in many traditional machine learning tasks, like image classification, recommender systems, text-to-speech, and so on. Nevertheless, there are still fundamentally problematic issues with these systems, that invite theoretical work on extensions of deep learning towards (traditional) probabilistic reasoning. This is a topic of some interest, which has lead to nice tools that can be used for implementation/evaluation, like Tensorflow Probability and Pyro (built on top of PyTorch). For typical research trends, see Part III of "The Deep Learning Book" by Goodfellow et al. The selection of interesting research question(s) in this area will depend on the students' interest.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Water management -- with TrønderEnergi

Introduction
TrønderEnergi runs a number of hydropower-stations, and one of the decision-problems they are confronted with is how to optimally use the water in the reservoir: The water level in the reservoir needs to stay within given bounds now and in the future, also considering the inflow of new water in the coming days. Add to this that the inflow is unknown, albeit prognoses do exist. Furthermore, the power-prices fluctuate, and one would obviously aim to maximize income by producing power at the points in the future where prices are at the highest. As for inflow, also future prices are unknown, but forecasts are available.

Problem description
Water management has traditionally been solved using convex optimization techniques, but TrønderEnergi has started looking into using AI methods to improve on the results. Several approaches can be considered:

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Your very own project!

The projects I have made available are simply proposals for you to consider. However, the best projects come from motivated students, and if you have your own ideas regarding a project that is related to

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Adversarial input attacks against obstacle detection

DNV GL as the classification society has devoted to research and exploration of the emerging technologies which are being applied to development of autonomous ships. Currently, DNV GL Group Technology and Research is actively developing a conceptual autonomous ship called ReVolt. ReVolt is dedicated as a platform for exploring and experimenting how to integrate novel technologies and equipment into autonomous ships to achieve the required level of ocean autonomy.
The most crucial and novel part of ReVolt is the navigation system which integrates both matured and novel maritime technologies, various types of sensors and machine learning technologies to achieve different levels of ship navigation. It has been widely recognized that cyber-physical systems highly depending on various types of sensors and complex algorithms to operate are both vulnerable to physical harm and subjected to cyberattacks. Therefore, we need to fully assess the potential cyber risks to assure that ReVolt can safely operate and complete its missions. Based on the current implementation of Revolt, we have identified several tasks to be carried to help DNV GL understand the feasibility of potential cyberattacks against the navigation system of ReVolt. Consequently, such learning and experience will help DNV GL define new class rules for certifying autonomous ships and develop Verification & Validation (V&V) requirements for autonomous ships.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Autofix feature of software security vulerability detection IDE plugins


Most of the current software security practices are to test the software using penetration testing at the very late stage of software development. As developers are not well trained to develop secure software, or their software security knowledge is not update, developers introduce software vulnerabilities when writing code.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

DevOps for AI

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Malicious AI

AI and ML are widely used for pattern recognition, intrusion detection, and malware classification, therefore offering promising solutions in cyber defense. However, our recent studies [1][2] show that an attacker can leverage AI technologies maliciously to create new cyberattacks or optimize existing cyberattacks. Such attacks can evade existing security detection mechanisms and augment attacking capabilities. This project aims to study how to identify, analyze, and mitigate AI-based security attacks. The tasks include:

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Privacy of blockchain-based supply chain system

The study will be part of the PaaSforChain (Platform as Service Technologies for High-performance Blockchain-based Supply Chain Management Systems) project, hosted by NTNU IDI. The PaaSforChain project is a Chinese-Norwegian Collaborative Project on Digitalisation financed by the Research Council of Norway and the Ministry of Science and Technology of the People’s Republic of China. The PaaSforChain project objectives include studying core technological, ethical, social, and cultural issues of using blockchain and other related IT technologies to develop blockchain-based supply chain applications, especially for the seafood industry and studying core IT technologies, such as architecture, security, privacy, performance, and scalability of developing a supply chain domain-specific PaaS platform.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Secure AI

As AI and ML is increasingly being used in a variety of security applications, like pattern recognition, intrusion detection and malware classification, vulnerabilities within these applications can be very impactful in the field of security. Exploring these vulnerabilities, and generate effective defensive strategies is therefore of high importance.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Security of blockchain-based supply chain system

The study will be part of the PaaSforChain (Platform as Service Technologies for High-performance Blockchain-based Supply Chain Management Systems) project, hosted by NTNU IDI. The PaaSforChain project is a Chinese-Norwegian Collaborative Project on Digitalisation financed by the Research Council of Norway and the Ministry of Science and Technology of the People’s Republic of China. The PaaSforChain project objectives include studying core technological, ethical, social, and cultural issues of using blockchain and other related IT technologies to develop blockchain-based supply chain applications, especially for the seafood industry and studying core IT technologies, such as architecture, security, privacy, performance, and scalability of developing a supply chain domain-specific PaaS platform.

Status: Valgbart     Egnet for: En student     Lenke: plink

AI-agents trained by Deep RL in simulated environments

Training of AI agents in virtual environments using deep reinforcement learning.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

AI-lab master's thesis pitches from "externals"

AI-lab master's thesis pitches related to Visual Intelligence (Computer Vision based on DL/ML/AI).

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Anonymization and Image Inpainting

Various topics related to anonymization and Image Inpainting and DeepPrivacy:

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Autonomous vehicles (AVs), AI/ML/DL and Computer Vision (CV)

Interested in AVs, AI and CV?

Have a look here

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Digital Twins (DTs), Wearable sensors and Medical Imaging++

Instersted in Digital Twins (DTs), Wearable sensors and Medical Imaging?

Have a look here

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

NAP-lab, Autonomous Vehicles (AVs) and Autonomous Driving (AD)

Using our new AV platform for AD.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

NVIDIA Clara Imaging / MONAI for Radiology and Pathology

Work in progress

NVIDIA Clara Imaging

Pre-trained models and transfer learning

AI-Assisted Annotation

Federated learning

etc.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Omniverse, Content generation, Simulation and Digital Twins

Ultimate Visual Computing and AI project.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Virtual, Augmented and Mixed Reality for learning AI etc.

Interested in Virtual, Augmented and Mixed Reality (i.e. XR) related to learning (e.g. AI/ML/DL) and training (e.g. medical cases)

Have a look here

 

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Vision Transformers (for Visual Intelligence)

Since its introduction in 2017, Transformers have revolutionised NLP and completely taken over most sequence processing.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Eco-visualization

Sustainable energy consumption requires behavioral changes. One way of motivating and supporting change is to make consumers more aware of their actual energy use, and to put this consumption into context. At the research center ZEN (Zero-emission neighborhoods), one of the research aims is to explore ways of visualizing energy consumption, both at the individual level and at the neighborhood level. The project will involve building various prototypes for this purpose and test them out. The prototypes will be developed for one or more of the existing neighborhoods of ZEN and should be tested on actual residents of these neighborhoods. In focus is to maximise the use of energy produced in the actual neighborhood.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Automated Engineering (AI Lab Pitch)

Description

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Automatic Metadata Generation: Language Technology at the Interface of Spoken and Written Norwegian

  • Advisor at NTNU: Prof. Ole Jakob Mengshoel - ole.j.mengshoel@ntnu.no
  • Contact(s) at NRK: Egil Ljøstad - egil.ljostad@nrk.no; maja.wettmark - maja.wettmark@nrk.no

Background

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Evolutionary Algorithms with Applications

Background

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Improved Ambulance Response via Improved Placement of Resources (AI Lab Pitch)

Proposal for master thesis at Norwegian Open AI Lab, NTNU, from Norwegian National Advisory Unit for Prehospital Emergency Medicine (NAKOS), and Department of Emergency Medial Communication Centre (EMCC), Division of Prehospital Services, Oslo University Hospital.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Individualized Feedback Generation Based on the Skill Level and Progression in Language Training

  • Advisor at NTNU: Ole Jakob Mengshoel, ole.j.mengshoel@ntnu.no
  • Contact at Capeesh: Martin Sandberg, martin@capeesh.com

About the Company

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Machine Learning for Causal Analysis of Air Quality

This project focuses on machine learning for causal analysis with a focus on interventions and counterfactual analysis, with application to air quality in Trondheim and other municipalities. The idea is to develop a causal model that includes things like exhaust from vehicles, use of studs, wood burning, ship traffic, cleaning of roads, …, and their impact on air quality. In other words, the goal is to develop a causal model containing causes and effects related to air quality, perform machine learning from air quality and related data, and improve understanding and decision making related to air quality in many municipalities.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Machine Learning for Effective Ocean Data Analysis

  • Advisor at NTNU: Ole Jakob Mengshoel. ole.j.mengshoel@ntnu.no
  • Contact at SINTEF Ocean: Ute Brönner, ute.broenner@sintef.no (Fraunhofer from 07/2021 - 07/2022)

Background

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Machine Learning wih Applications

This projects focuses on machine learning and its application; in particular the areas of environment, health, medicine, and transportation are of interest. Beyond the machine learning and application dimensions, one or more of the factors safety, explainability and sustainability will typically be important.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Memetic and Multimethod Stochastic Optimization Algorithms

Project Background

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Next Generation Responsible Gaming Solutions Driven by AI

Norsk Tipping has the ambition to be world leading within responsible gaming. As part of the company’s mandate Norsk Tipping is required to provide and offer gaming activities in a safe and secure environment under public control with the aim of preventing the negative consequences of gambling. This project focuses on the use of and research on AI technologies for this purpose.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Reinforcement Learning for Black-box Safety Validation of Autonomous Marine Vessels

  • Advisor at NTNU: Prof. Ole Jakob Mengshoel - ole.j.mengshoel@ntnu.no
  • Contact at Zeabuz: Dr. Øyvind Smogeli -  oyvind.smogeli@zeabuz.com
  • Contact at NASA: Dr. Ritchie Lee

Backgroun

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Stochastic Local Search: Algorithms and Applications

In AI and ML, several methods rely on stochasticity or randomization: mutation and crossover in evolutionary algorithms; dropout and stochastic gradient descent in deep learning; stochasticity in stochastic local search (SLS); and randomization in systematic search. SLS algorithms, which we study here, are competitive in solving computationally hard problems such as satisfiability (SAT), sparse signal recovery, scheduling, and most probable explanations in Bayesian networks (BNs). Essentially, SLS algorithms are greedy optimizers that also make random moves in order to avoid getting trapped in local but non-global optima.  Further, SLS algorithms are interesting in that they can studied formally, for example by means of Markov chains.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

3D oriented data augmentation for Deep Object Detection

As it is well known, recent deep learning methods allow unprecedented performance for image classification, semantic segmentation, and object detection tasks. When building object detectors for marine applications, it is particularly important to have training sets which do not only contain a sufficient number of relevant object classes (ships, piers, buoys, ...), but also show the different variations in lighting conditions, and weather conditions (fog, rain, snow, ...). As it is difficult, if practically possible at all, to record a sufficient number of such images or videos for training object detector networks, this project aims at artificially generating these variations of image conditions, in particular different illumination and strongly different weather conditions.
Image augmentation is a well-known technique for synthetically expanding limited or imbalanced training datasets used in computer vision tasks to improve generalization performance and to avoid overfitting. Likewise, augmentations can be used for testing network generalization performance, by imposing varying augmentations, e.g. two-dimensional augmentations such as shifts, rotations, etc., as well as natural phenomena, e.g. snow, fog, lighting conditions, etc., or even adding synthetic objects to the scene. The former is straightforward using simple transformations in two dimensions, whereas the latter should imitate realistic, and highly stochastic, scenarios that in reality depends on the depth information of the scene. Nonetheless, this is often done by basic image manipulations imposed uniformly over the whole scene without considering depth information and may result in unrealistic augmentations not representing reality and could possibly produce erroneous testing performance. As an example, fog or rain may be observed only at a certain distance and/or area and affects only specific far-away objects, rather than all objects in the scene. The proposed research topic is therefore to investigate techniques for extracting depth information from images or video and taking advantage of this information to apply augmentations of natural phenomena in 3-Dimensions.
For the extraction of this 3D information, one can use recent methods of single image depth estimation by deep neural networks, or "motion stereo" from a moving camera (in case of video sequences). Both approaches can also be combined.
Therefore, the first partial goal of this project is to explore methods for depth estimation for given training images or videos.
The second goal is to use the extracted estimate of the depth structure to simulate the desired weather and illumination situation, such as fog, rain, or snow. Again, this can be done by "classical" physics-based computer graphics modeling, and/or by suitably trained deeep neural networks.
Another important aspect is to benchmark 2D and 3D synthetic augmentations against images containing actual natural phenomena’s, i.e. network performance on synthetic scenes against true scenes.
This project will be performed with the Norwegian company DNV GL which is interested in quality assurance of marine detection systems, and will also be related to the marine multi-sensro simulator project currently pursued at NTNU.
Students who are interested in this project are advised to address Prof. Rudolf Mester at IDI for more details. The co-supervisors will come from DNV GL and probably also from the Cybernetics Department (ITK). 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

AI for Mobile Systems: Let us define a project after your own interests

Besides the more or less fixed topics which are found in the list, there is always the possibility to define a project according to your own interests, as long as it is scientifically solid or a real engineering challenge.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Building a simulation framework for autonomous driving

This project aims at creating a game engine based simulation environment for autonomous driving with a strong emphasis on physically realistic car dynamics, realistic road networks (based on OpenStreetMap), allowing for massive multi-agent simulations.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Building an intelligent mobile robot for cooperative navigation or communication

This project aims at building a mid scale mobile robot, which could use the same hardware platform as the model race cars used in the MIT model race car project https://mit-racecar.github.io

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Building Ship Models for Simulators using Computer Vision and AI

# Building digital models of ships from video sequences

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Collision Resilient Navigation for Aerial Robots in Confined Environments

In this research-intensive project we investigate the potential of using reinforcement learning to enable a collision-tolerant flying robot to navigate safely in its environment. Starting from an existing collision-tolerant vehicle design you are on one hand encouraged to modify but also to to model it so as to enable the use of reinforcement learning in simulation. You will be provided a planner that allows collision-free navigation and exploration in unknown environments as a reference. Then you are tasked to design a Reinforcement Learning (RL) agent acting in the continuous control space so as to optimally guide the robot such that it navigates safely in its environment. Given its collision-tolerance, “safely” may not necessarily mean with zero collisions and you are challenged with the idea that if a robotic embodiment allows some more tolerance then the navigation function can possibly become more agile and efficient, while remaining safe. The vast amount of training should happen in simulation and combined with selective real-life data (smooth sim-to-real transfer). The derived RL agent is to be implemented onboard the actual robotic system and experiments in safe navigation inside narrow corridors are to be conducted to evaluate its performance. It is noted that the proposed work is aligned with an ongoing project for the DARPA Subterranean Challenge (namely research activities of Team CERBERUS): https://subtchallenge.com/

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Computer Vision in Bad Weather conditions / Bad Visibility

This project is a cooperation with a leading industrial partner that is one major player in the field of autonomous driving.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Convolutional neural network based object tracking in drone-based inspection

This is one of three related topics originating from a cooperation between DNV GL and NTNU, addressing the task of inspecting ship tanks for cracks and similar faults, using a drone equipped with a camera.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Detecting cracks from a stack of motion compensated drone images

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Egomotion estimation and map building for the Revolve Autonomous Race Car

This project is related to the Revolve student project aiming at building / improving an Autonomous Race Car that can participate in the international Formula Student competition. Revolve is participating to this yearly competition already for a number of years. Each year, a new version of this race car is developed, and of course it is a fundamental goal of the overall Revolve project to achieve a good rank in the annual competition.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Environment perception for a mid scale robotic model car

The perception of the environment, the determination of drivable area, and the detection and classification of obstacles is a central task when it comes to building an intelligent mobile robot.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Environment Perception for Underwater Robots

This project is offered in the context of a new upstarting research project which is performed as a cooperation between the cybernetics department (ITK) and computer science (IDI). The overall project AROS deals with the task to provide autonomy to the snake underwater robots developed by ITK (Prof. Kristin Y. Pettersen and colleagues). It is funded by the Norwegian Research Council.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Multi-Camera Traffic Surveillance on Urban Crossings

This project aims at developing methods for processing synchronized video streams from multiple cameras mounted at large urban crossings. The approach to be taken should allow for a setup of such a system with a minimum of human effort, applying methods for automatically finding out about the image-to-image relationships. An example of a learning-based approach allowing this is the work of C.Conrad et al. [ https://ieeexplore.ieee.org/abstract/document/5981689 ]

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Multi-Modal Thermal-Visual-Inertial Odometry in Smoke and Dust-filled Environments

In this research-intensive project you are tasked with the goal of designing a multi-modal sensor fusion framework that exploits both visible-light and thermal camera data, as well as inertial measurements (i.e., accelerometer and gyroscope readings). The purpose of the approach is to enable resilient localization in environments that are not only GPS-denied but further degraded by the presence of dust and smoke. To achieve this goal you are tasked to utilize some type of fixed-lag estimator or possibly do extensive use of pose graph optimization techniques in combination with appropriate design of feature extraction, matching and tracking front-ends. For the latter you are encouraged to investigate a wide variety of approaches, including direct- and semi-direct methods. The designed method will be tested onboard a flying robot integrating a time synchronized visual-thermal-inertial sensing module. Field experiments will be conducted in underground environments filled with dust or smoke obscurants. It is noted that the proposed work is aligned with an ongoing project for the DARPA Subterranean Challenge (namely research activities of Team CERBERUS): https://subtchallenge.com/

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Object detection and multiple objects tracking in drone based inspection

This is one of three related topics originating from a cooperation between DNV GL and NTNU, addressing the task of inspecting ship tanks for cracks and similar faults, using a drone equipped with a camera.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Quantifying the reliability of results from deep neural networks

This project attempts to answer the question how reliable the mostly stunning results of deep neural networks used for classification (what kind of object is this?) or estimation (how distant is this pedestrian from my car's camera?) are.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Robust, fast and lightweight environment perception for next generation autonomous race cars

This project is related to the Revolve student project aiming at building an autonomous race car. Its purpose is to explore methods of furthering the utility of the stereo camera system used in the current vehicle, either as a fusion of LiDaR and camera, or as a pure stereo camera configuration.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Self-learning visual perception for a mobile robot

In this project, a motion perception system for a robot is to be developed. The main idea is that ego-motion of a robot can be learned from the visual input in combination with data from an inertial measurement uni (IMU) which is able to sense translational acceleration and rotational velocity. This is (hypothetically) the same way how humans and animals learn to perceive their motion, by combining data from visual input and the inner ear (sense of balance).

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Structured Deep Object Detection

The most powerful modern visual object detection approaches today are based on deep learning. They usually require large sets of labeled training data.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Visual Simultaneous Localization and Mapping (VSLAM) for an Underwater Snake Robot

Visual SLAM is a term for a set of methods and algorithms that a) determine the motion of a camera (or a set of cameras) through an environment and b) determine the geometrical shape of that environment. vSLAM often builds on detecting “prominent points” in the images, and tracking them through the sequence. If a sufficient number of such points are tracked between two images, the relative pose (=translation and rotation) of the camera can be estimated. As any measurement in images is afflicted by errors, both these pose estimates as well as the estimated 3D positions of the observed image points are uncertain, and the estimation of the complete camera trajectory as well as the scene model “stitched together” from many views needs to be input data to a huge optimization problem.
In AROS, we have access to both real video footage from underwater missions, as well as a realistic simulation environment which is able to generate video sequences where the motion and the 3D geometry are precisely known (‘ground truth’). The student project is integrated into our design and development process for a vSLAM system which is specifically tuned to be able with the substantial problems of underwater video material: limited visibility due to turbid water, bad illumination which is also moving with the robot vehicle, disturbances by plankton, dirt, and small fish, and many more. Which part of the vSLAM development is determined to be the focus area of the student project is subject to negotiation; the intention is to let the students experiment with novel approaches proposed in the recent literature, some of them focusing on geometric models and statistical estimation theory, others on machine learning. So we are able to adapt the topic largely to the background knowledge the student(s) already have, and their interest into different relevant research fields, such as e.g. state estimation, optimization, object detection and tracking, machine learning and deep learning.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Performance Study of Geometry Encodings for Finite Volume Methods in Computational Fluid Dynamics

* Oppgaven er forbeholdt for Jenny Manne

Status: Valgbart     Egnet for: En student     Lenke: plink

Artificial Intelligence in Business: Uncovering challenges and obstacles of adoption.

The last few years have seen an explosion in interest regarding the use of Artificial Intelligence and much talk about the potential business value. Nevertheless, there is significantly less talk about the challenge's organizations will face when implementing such solutions and how they should overcome these obstacles. Inhibiting factors are not only of a technological nature but also include organizational and human factors. This project will involve collecting and analyzing data in collaboration with the researchers from the Big Data Observatory (https://www.observatory.no).

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Artificial Intelligence in the Healthcare sector

Advanced forms of analytics and aritificlai intelligence are becoming increasingly deployed to support the work of healthcare workers. Medical doctors, nurses, and administrative staff either use, or are aided by sophisticated technologies which are posed to radically change the nature of their work. For example, radiologists now rely increasingly more on machine learning techniques to and other applications of AI to diagnose patients, while a lot of procedural and repeptive tasks are being done by machines. The objective of this project is to understand how the nature of work for health practitioners is changing, and what positive and negative consequences they experience.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Big Data Analytics: Challenges and obstacles in deployments

Over the last few years we have seen more companies trying to deploy big data analytics to outperform competition. Nevertheless, there is ongoing debate about whether such investments do indeed create value if so how this value can be captured and what are the main challenges in doing this. The objective of this master thesis is to perform a qualitative study through interviews and focus groups on companies in Norway and examine the ways in which they are applying big data analytics to create business value, the stages they go through in implementation and the obstacles they face. The project is in cooperation with the Big Data Observatory (https://www.observatory.no), during which you will learn how to develop research methods and analyze qualitative data.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Deciding what to purchase: How do we consume information on social commerce platforms like Amazon?

An increasing amount of purchases are done through social commerce platforms such as Amazon. These platforms differ from conventional e-shops in that they contain a lot of information that is produced by consumers themselves. This creates an interesting mix, where potential buyers have to navigate through information that is produced by marketers and consumers and based on this make assessment of which products to buy. Research has suggested that the types of products we buy also influences how we process information. This project will focus on examining consumer behaviors on popular social commerce website such as Amazon. It will include the use of eye-tracking techniques with an experienced researcher and involve the design and execution of studies with participants and subsequent interviews with them.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Exploring the black box of AI Governance

AI governance is a notion that is often attributed to a range of different practices and processes. From establishing a process of developing AI applications, ensuring that quality outcomes are achieved, and to decising the role and responsibilities of stakeholders. AI governance now plays an important part related to the business value that AI can deliver, and to ensuring that projects comply with ethical and regulatory frameworks. This project will seek to understand how organizations develop AI governance practices, what aspects they take intio accoutn when doing so, how they deploy them, and what the outcomes of them are at the business and project performance levels.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Human-AI interaction and decision-making

As a increasing amount of organizational decision-making is based on insght generated by sophisticated algorithms, there is a need to understand how key stakeholders interact with interfaces that aid such decisions. From insight to action requires that generated insight is presented and designed in such a way so that it can be readily "consumed" byt decision-makers, and that it satisfies informational requirements without overloading the recipient. Such design must incorporate several aspects, such as ifnroamtion presentation, detail, interface design, interactivity and others. If AI-generated insight is to reach its potential, it is important that we understand how to develop such interfaces with human decision-makers under different contextual and time-related requirements

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Human-Machine Symbiosis: How does decision-making change in the age of Big Data Analytics?

The nature of decision-making is changing drastically, both in personal lives and in the business sphere. An increasing amount of decisions are now based on insight that is generated through analytics. Despite this, often individuals are faced with cognitive-overload, conflicted views, or biases that result in non-adoption of insight. This project will be done in collaboration with the Big Data Observatory (https://www.observatory.no) and involve designing a study protocol and collecting and analyzing neurophysiological data (eye-tracking and electroencephalography) from study participants. This will be done with the help of an expert in such tools.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Organizational decision-making in the age of AI

Artificial Intelligence is now being used at an increasing rate to augment or automate organizational decision-making. From processes such as performing credit checks on customers of banks, aiding in forecasting of future events, and automating manual and repetitive tasks, AI is introducing a new way of making decisions for organizations. The purpose of this project is to examine through empirical methods the effects and processes of transition to AI-based decision-making structures.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Responsible AI in organizations

The notion of responsible AI entails a large range of aspects regarding how AI applications are developed, utilized, and monitored throuhgout their lifecycle. The purpose of this project is to explore what responsbile AI means for organizations, which processes and structures they are establishing in order to attain set indicators of responsible AI, as well as what are the organizational impacts of it. Does adopting responsbiel AI result in any organizational gains? Does it influence how customers/citizens perceive the organization, or is it restricting what they can do with novel technologies?

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Social commerce: What makes consumers browse, buy and spread word-of-mouth

Social media platforms have become more and more commerce focused. Now platforms such as facebook, Instagram, snapchat and twitter feature context from sellers and also enable consumers to buy directly through them. Yet, behavior of users and what influences their decisions to buy or spread interesting context is not well understood. The objective of this study is to examine how design features and usability aspects influence behavior on such mediums. The project will include collecting and analyzing data through quantitative research methods such as survey design.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

The potential of Artificial Intelligence for public administration

While there has been a long discussion about the potential of using Artificial Intelligence in private organizations, now more and more public organizations are implementing solutions to support their operations. From uses for fraud detection, chatbots, autonomous vehicles, or infrastructure monitoring, AI is gaining ground in applications for public administration. This project will be done in connection with SINTEF Digital and will involve data collection, analysis and reporting. The aim is to find out what is the status of AI adoption, what are the potential interesting uses, and what is the value that is realized.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

The Strategic Value of Big Data Analytics

Today more and more companies are using big data analytics to support or drive their business strategies. Yet, there is ongoing debate about whether such investments do indeed create value if so how this value can be captured. The objective of this master thesis is to perform a quantitative study on companies in Norway and examine the ways in which they are applying big data analytics to create business value. The project is in cooperation with the Big Data Observatory (https://www.observatory.no), during which you will learn how to develop research methods and analyze quantitative data.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

What are the skills of the future: A look at how Artificial Intelligence and Analytics change the competences of computer science graduates

Over the last years the emergence of key technologies such as big data analytics and artificial intelligence have given rise to a completely new set of skills that are needed in private and public organizations. With IT gaining an increasingly central part in the shaping of business strategies, it is important that study curricula follow these requirements and provide graduates that fit the needs of organizations. This project will be run in collaboration with the Big Data Observatory (https://www.observatory.no) and involve collecting data through focus groups and surveys with key representatives. The output will involve a detailed look at what skills are necessary and how they can be addressed by educational institutions.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

What is the Business Value of Artificial Intelligence?

While there has been a lot of focus on the technical aspects related to artificial intelligence, recent years have seen a growing discussion about what the application of AI could be for private and public organizations. The objective of this master thesis project is to examine the readiness of private and public organizations to adopt AI, and the value they have derived from such investments. This project will involve collecting and analyzing data in collaboration with the researchers from the Big Data Observatory (https://www.observatory.no). It is an exciting opportunity to see how organizations are planning to use AI and what steps they need to take to adopt such technologies.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Data science - in practice

Data-driven data science is attracting a lot of interest.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Digitalisering i olje og gass: IoT

Stadig mer av oljeutvinningen på norsk sokkel foregår på svært utilgjengelige felt på store havdyp. Samtidig benyttes stadig oftere ubemannete subsea anlegg som ligger på havbunnen men fjernestyres fra land eller nærliggende rigg. Sensorbasert informasjon (bl.a. temperatur, trykk, seismikk, magnetiske og strålingsegenskaper til bergformasjonen) spiller stadig større rolle i alle fasene: leting ("exploration"), boring ("drilling"), logging og produksjon.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Health Information systems in developing countries

Developing countries have limited resources for healthcare delivery hence need to make the most of resources available.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Open source: læring og kunnskapsdeling

Open source ('åpen kildekode') utvikling får økende oppmerksomhet, ikke minst pga. det faktum at mye av den mest interessante teknologien idag er utviklet (mer eller mindre) open source.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Samarbeide og kommunikasjon i helsesektoren

Helsesektoren er fragmentert geografisk, institusjonelt og gjennnom spesialiseringer. Sektoren har utfordringer knyttet til at geografiske, institusjonelle og profesjonelle grenser hindrer effektiv kommunikasjon som er en forutsetning for god pasientbehandling.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Detection of security vulnerabilities in source code using machine learning

Writing secure software is a challenging task and with the landscape of known security vulnerabilities changing almost daily, this task is practically unending. There has been some effort in recent years to use machine learning to automatically scan software repositories during the development lifecycle [1, 2, 3].

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Discovery of program control flow in binary programs from unknown instruction set architectures

Reverse engineering (RE) is the process of discovering features and functionality of a hardware or software system. RE of software is applied where the original source code for a program is missing, proprietary, or otherwise unavailable. Motivation for RE ranges from extending support of legacy software to discovery of security vulnerabilities to creating open source alternatives to proprietary software.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Machine learning for reverse engineering of unknown instruction set architectures and custom virtual machines

Reverse engineering (RE) is the process of discovering features and functionality of a hardware or software system. RE of software is applied where the original source code for a program is missing, proprietary, or otherwise unavailable. Motivation for RE ranges from extending support of legacy software to discovery of security vulnerabilities to creating open source alternatives to proprietary software.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Multi-Image Bilirubin Estimation for detection of neonatal jaundice

Neonatal jaundice
An estimated 114000 newborns die and 179000 newborns get permanent brain damage every year due to the lack of proper management of neonatal jaundice [3]. Three-quarters of these cases occur in Sub-Saharan Africa and South Asia, showing that low-income countries lack the tools and processes needed to deal with the condition adequately. Contrary, there are very few cases of deaths and brain damages occurring in high-income countries, roughly 1 in every 100000 births. Still, in high-income countries vast resources are spent dealing with the condition, with jaundice being the number one reason for readmissions to the hospital after birth [4].

Multi-Image Bilirubin Estimation
The objective is to explore whether a higher accuracy of bilirubin estimates can be achieved through methods capable of “digesting” multiple images from one newborn at once, rather than performing bilirubin estimation for a single image.

We will provide you with a dataset consisting of sets of skin images from newborns and targets (bilirubin concentrations). Your task is to explore, design and implement methods capable of estimating bilirubin concentration of a newborn using jointly multiple images of skin, or features extracted from them.

This is a continuation of ongoing research.

Picterus
Picterus [1] is a growing start-up bringing biophysics and machine learning to healthcare with the ambition to provide accurate and low-cost solutions for jaundice diagnosis. Picterus has developed a smartphone app as a novel approach for neonatal (newborn) jaundice diagnostic and monitoring [2]. Particularly, Picterus is using computer vision to estimate bilirubin levels, the chemical compound causing jaundice at high concentrations.

Dataset
We own a unique dataset of newborns’ skin images coupled with bilirubin concentrations that Picterus has collected across multiple continents.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Unsupervised Detection of Anomalies in Skin Images

Neonatal jaundice
An estimated 114000 newborns die and 179000 newborns get permanent brain damage every year due to the lack of proper management of neonatal jaundice [3]. Three-quarters of these cases occur in Sub-Saharan Africa and South Asia, showing that low-income countries lack the tools and processes needed to deal with the condition adequately. Contrary, there are very few cases of deaths and brain damages occurring in high-income countries, roughly 1 in every 100000 births. Still, in high-income countries vast resources are spent dealing with the condition, with jaundice being the number one reason for readmissions to the hospital after birth [4].

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Weakly Supervised Approach for Detecting Important Pixels

Neonatal jaundice
An estimated 114000 newborns die and 179000 newborns get permanent brain damage every year due to the lack of proper management of neonatal jaundice [3]. Three-quarters of these cases occur in Sub-Saharan Africa and South Asia, showing that low-income countries lack the tools and processes needed to deal with the condition adequately. Contrary, there are very few cases of deaths and brain damages occurring in high-income countries, roughly 1 in every 100000 births. Still, in high-income countries vast resources are spent dealing with the condition, with jaundice being the number one reason for readmissions to the hospital after birth [4].

Weakly Supervised Approach for Detecting Important Pixels
We hypothesize that certain pixels or clusters of pixels of a skin image do not provide any predictive information and may confuse some types of estimation methods. For example, highlights or shadows are such types of pixel clusters that most likely do not contribute to an accurate estimation of bilirubin.

RQ: How can we learn without any labeled data (ground truth masks) a model that would be able to identify pixels that contribute to an accurate bilirubin estimation?

Here we have several promising and detailed hypotheses to test, but they may require a low level (mathematical) understanding of neural nets.

The potential contributions are at least threefold:

    1. Insights on which parts of the skin image carry the predictive information for bilirubin estimation, and whether this is dependent on the estimation method.
    2. Automated feature selection improving bilirubin estimation.
    3. Partial explainability of estimations.

Picterus
Picterus [1] is a growing start-up bringing biophysics and machine learning to healthcare with the ambition to provide accurate and low-cost solutions for jaundice diagnosis. Picterus has developed a smartphone app as a novel approach for neonatal (newborn) jaundice diagnostic and monitoring [2]. Particularly, Picterus is using computer vision to estimate bilirubin levels, the chemical compound causing jaundice at high concentrations.

Dataset
We own a unique dataset of newborns’ skin images coupled with bilirubin concentrations that Picterus has collected across multiple continents.
Requirements
    • A friendly relation with programming and machine learning.
    • Experience with Python and machine learning frameworks such as PyTorch or TensorFlow.
    • Basic knowledge of calculus, linear algebra and statistics are beneficial.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Climbing Mont Blanc Back-Ends and Energy Efficiency Analysis

Climbing Mont Blanc (CMB) is a system for training and competitions in energy efficient programming of small processors. CMB has been using a heterogeneous multicore (MPSoC, Exynos from Samsung) in more that five years. The chip is accessed via an Odroid XU3 board that has integrated energy monitors. These boards are no longer produced, and the energy measurements are not as precise as we would like. The goal of the proposed project is to be able to use one or two newer single board computers (SBC) as CMB back-end and the Lynsyn system to achieve more precise energy measurements. The project can be continued as a master thesis project in the spring semester.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Blockchain-based Data Marketplace

The European regulations about privacy, the GDPR, is changing people’s understanding and attitude towards data/information they own and those the others own. The ownership of data/information has always been a concern but with GDPR there will be more awareness as well as more obligations related to data collection and sharing.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Decentralized AI

Nowadays, big tech giants, such as Google, Facebook, Amazon, IBM and Microsoft, are dominating the AI market by offering cloud-based AI solutions and APIs. They are collecting user data in one place through free services and systems, analyze it for insights and resell to third parties, such as advertisement companies. This model is centralized AI, which is working fine now. But in the long-run it could lead to monopolization of the AI market. This could also cause unfair pricing, lack of transparency, interoperability, privacy issues and excluding smaller companies from AI innovation. Fortunately, there is the emergence of a decentralized AI market, born at the intersection of blockchain, on-device AI and edge computing/IoT.
In this project, we will investigate the possibility to build a proof-of-concept of a decentralized AI application through blockchain, such as Ethereum. AI agents will train and learn models from their own data. The decentralized AI application to be developed in this project then can combine multiple algorithms/models (developed by different agents) performing different sub-tasks. One of possible applications is Decentralized Autonomous Cars. 1 - 2 students.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Deep Learning to combat with micro-plastic pollution

An estimated 275 million tonnes of plastic waste was produced on a global scale in 2010, with 8 million of those tonnes being introduced to the oceans - about 3% of global annual plastics waste. Once the plastic reaches the oceans, it is broken down into smaller particles(micro-plastic) by being exposed to ultra violet (UV) radiation and mechanical abrasion from wave actions [1].The quantity of plastic waste floating at the ocean surface in 2013 was estimated to be approximately 269,000 tonnes (small macro- to micro-plastic), this estimate does not include plastic in-depth or at the seafloor). The plastic debris can affect the wildlife in multiple ways, such as entanglement- entrapping, encircling, or constricting,ingestion- accidental ingestion or ingestion of prey containing plastic, and interaction- being in contact with plastic debris [1].It is therefore important to be able to detect and collect the plastic waste in nature,before it reaches the oceans. Once plastic waste has reaches a micro-stadium, it is near impossible to collect it and remove it from the water. An analysis on deep sea locations(range from 1176 to 4843m) showed that there was an average abundance of 1 micro-plastic per 25cm3(particle sizes ranging from 75 to 161μm) [2]

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

General Data Protection Regulation (GDPR) and opportunities it brings: analysis of my all-data and generating insights

After enforcement of GDPR (General Data Protection Regulation) (May 2018), all companies and institutions collecting data about individuals are obliged to deliver to people the data they collected about them (e.g., whatever facebook, google, amazon, insurance companies etc collects about me shall deliver the data they collected about me when I asked for it). GDPR will give a chance for people to look into her/his data stored by those companies.
So, what people can do with so much (and rich) data about themselves? It would be super hard for them to analyze, extract insights and even look into the raw data downloaded from Facebook, Google, etc..

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Performance and scalability of public blockchains

This project concerns the problem of performance metrics and scalability of public blockchain systems. The student should conduct a systematic literature review and establish what performance and scalability characteristics and metrics are relevant in the domain of public blockchains. What are the current state-of-the-art solutions and algorithms, what is the current literature coverage and what are the main challenges. The student is expected to provide both, qualitative and quantitative analysis of the existing body of literature related to the area.

[1] Scherer, Mattias. "Performance and scalability of blockchain networks and smart contracts." (2017).

 

Status: Tildelt     Egnet for: En student     Lenke: plink

[ NorwAI] Mining dynamic graphs

For many important applications, data is presented as graphs, with dynamic relationships between nodes. Examples include graphs representing the power grid, financial transaction relationships, social networks, and transportation networks. The nodes in the graphs can also have attached dynamic data, for example time series, adding further challenges. The aim of this project is to 1) study previous work related to this topic, 2) identify interesting and useful data mining operations to be performed on dynamic graphs, 3) develop algorithms/indexes for efficient execution of one or more of these operations, and 4) evaluate these on a large dataset.

Prerequisites: TDT4150 or TDT4225, good programming skills and interest in algorithms.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NorwAI] Next best offer/identifying “twins” based on transaction data

This project is in cooperation with, and with co-supervisor from, Sparebank 1 SMN.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NorwAI] Querying dynamic graphs

For many important applications, data is presented as graphs, with dynamic relationships between nodes. Examples include graphs representing the power grid, financial transaction relationships, social networks, and transportation networks. The nodes in the graphs can also have attached dynamic data, for example time series, adding further challenges. The aim of this project is to 1) study previous work related to this topic, 2) identify interesting and useful query operations to be performed on dynamic graphs, 3) develop algorithms/indexes for efficient execution of one or more of these operations, and 4) evaluate these on a large dataset.

Prerequisites: TDT4150 or TDT4225, good programming skills and interest in algorithms.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NorwAI] System for efficient storage and retrieval of dynamic graphs

For many important applications, data is presented as graphs, with dynamic relationships between nodes. Examples include graphs representing the power grid, financial transaction relationships, social networks, and transportation networks. The nodes in the graphs can also have attached dynamic data, for example time series, adding further challenges. The aim of this project is to design and implement a system for efficient storage and retrieval of such data, based on a NoSQL database system, in more detail: 1) study previous work and systems related to this topic, 2) design and implement a prototype system, and 3) evaluate its performance.

Prerequisites: TDT4225 (preferably) and good programming skills.
 

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Mining of intertransaction association rules on Spark

In this project, the aim is to study mining of intertransaction association rules [1] and how to perform this efficiently and scalable on Spark.

Prerequisites: TDT4305, good programming skills and interest in algorithms.

[1] https://ieeexplore.ieee.org/abstract/document/1161581

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Semantic Trajectory Similarity Join on GPU

Semantic trajectories are trajectories (sequences of locations) with associated text/keywords. A similarity join query of such trajectories is the problem of finding pairs of trajectories that are sufficiently similar to be a "match". The task of this project is to study such joins, and how to execute this query efficiently on GPUs.

Prerequisites: Preferably TDT4150 and/or TDT4225, good programming skills, competence in C++ (or strong motivation to learn C++) and interest in algorithms.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Semantic Trajectory Similarity Join on Spark

Semantic trajectories are trajectories (sequence of locations) with associated text/keywords. A similarity join query of such trajectories is the problem of finding pairs of trajectories that are sufficiently similar to be a "match". The task of this project is to study such joins, and how to execute this query efficiently on Spark.

This project is most suited for group of two students.

Prerequisites: Preferably TDT4150 and/or TDT4225, good programming skills, and interest in algorithms.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Top-k spatio-textual join on GPU

In this project, the aim is to study top-k spatial join [1] with additional textual attributes [2,3,4], and how to execute this query efficiently on GPUs. 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Gamification and visual analytics in apps to support everyday learning activities

Supervisors: Sofia Papavlasopoulou, Kshitij Sharma and Michail Giannakos
Place: LCI Lab: https://lci.idi.ntnu.no/, Trondheim
Suitable for: One or Two students
 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

AI applications for social good (AI4SG)

Supervisors: Ilias Pappas, Letizia Jaccheri

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

The role of employee-driven innovation within private and/or public organizations

Supervisors: Leif Erik Opland, Ilias Pappas, Letizia Jaccheri

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Coordination challenges for long-term environmental research infrastructures in Europe

Through strategic programs such as ESFRI (European Strategy Forum on Research Infrastructures), the European Union is increasingly organizing long-term environmental research around networks of transnational Research Infrastructures (RI). One example is the European Long-Term Ecosystem Research (eLTER) RI, encompassing 27 countries and a network of sites researching on different environmental aspects. The EU regulations are also adapted and adopted in Norway, influencing the national environmental research strategy.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

IoT and data management in long-term environmental monitoring

Environmental research is a crucial activity to assess the health of the marine, terrestrial and atmospheric environment over time – for example to detect climate change. Monitoring is typically arranged in networks of research stations, each focused on a few key aspects of the natural environment (e.g., fresh or marine waters, agricultural or alpine areas). The main goal of long-term environmental research is the collection, sharing, and maintenance of the environmental datasets over the long term, to support future (re)use and interpretation.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Crowdsourcing for Collaborative Competence Development at the Workplace

This task will include prototyping concepts and ideas to support a group of people in organisations to collaboratively develop their competences. Concepts such as crowdsourcing content, co-creation and Community of Practice will be central. The work will be conducted within a Norwegian research project called SAMSPILL: https://www.sintef.no/prosjekter/2020/samspill-samskapt-utvikling-av-smarte-laringsspill-i-finansbransjen/

Status: Valgbart     Egnet for: En student     Lenke: plink

ICT and data architecture and Smart cities

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Mobile Application for Global Employability Skills

This work will explore the concepts of skills and competences that are relevant in a work context and design and evaluate a mobile application to help students document their skills and competences. The tasks will include:
• A literature review of relevant concepts, relevant design and evaluation concepts.
• Explore and develop concepts and prototypes of front-end design.
• Conduct evaluations and iterative improvements, with focus on user interface and usability.
This work will be conducted within the European ERASMUS+ project GES App.
Prerequisites – prior knowledge or interest in interaction design and app development will be useful.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Modelling Smart and Sustainable Cities

ICT plays an important role as enabling technologies in the field of smart and sustainable cities. The Smart cities concept often takes on a limited view of a city and tends to focus on one or few aspects of a city. In this project, we would like to explore the possibilities of modelling a city by looking at it from a holistic way. We would use the ideas of conceptual modelling and enterprise architectures to understand a city. We would also like to explore the city as a complex system and use complex systems modelling approach to simulate how a city evolves. The outcome of the project will be a model of a city. The tasks include:
- Literature review of enterprise architectures for a smart and sustainable city.
- Literature review of how to model a city.
- Design and develop a model of a city as a complex system.
The work could be extended to a Masters project where the model would be further enhanced and evaluated.

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Smart City Enterprise Architecture for Service-based Ecosystems

In a complex enterprise, such as city Enterprise Architecture, there are many challenges and opportunities for ICT and ICT could support the city in many ways. With new technological developments such as the easy access to Open Data, storage and analytics services, the ICT infrastructure is no longer in-house. The Enterprise Architecture for a city has been described as a Service-based ecosystem. This project will explore new and emerging ideas for Enterprise Architecture for Smart Cities and the ICT solutions to realise the concept of Service-based ecosystems. Some of the technological solutions include API management, Open data models, Open data, etc. The tasks include the following:
• Literature review of Service-based ecosystem, Enterprise Architecture for Smart Cities.
• Overview of relevant approaches, technologies and solutions, Open data Models, Standards.
• Implementation of a simple Enterprise Architecture prototype.
• Evaluation of the work.
This work will be conducted within the European project +CityXchange (https://cityxchange.eu/) and will most likely collaborate with some of the partners in the project, e.g. Trondheim Kommune.
Prerequisites: prior knowledge about knowledge Enterprise Architecture and Smart Cities and TDT4252 course would be beneficial.

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NorwAI] Mobility Data with Telenor Research

This project is connected to the Norwegian Research Center for AI Innovation (NorwAI) and the Norwegian Open AI Lab (NAIL). 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

AI methods for CP Diagnosis

Brain injury in newborns accounts for more than 9 million years lived with disability worldwide. The functional consequences of early brain injury in newborns are commonly seen months or even years after birth, delaying therapeutic interventions and leaving families with uncertainty about their child’s health status for months and years.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NAIL Master Thesis pitch 2021/2022 collaboration with International team INEOS] - Data Driven Generative Model for Cyclist Performance Simulation

[more information here]

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NAIL Master Thesis pitch 2021/2022 Collaboration with Refinitiv AS] Deep Learning Models for detecting wind turbines’ curtailment

[More information here]

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NAIL Master Thesis pitch 2021/2022] - Approximating computational fluid dynamics (CFD) simulations using Deep Neural Network

More details here. (thesis in collaboration with NablaFLow)

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NAIL Master Thesis pitch 2021/2022] AI-facilitated performance-development in endurance sports

The thesis will be carried out as a collaboration between NAIL and Centre for Elite Sports Research (SenTIF) at NTNU. Research at the SenTIF revolves around key issues of the Norwegian elite sports. SenTIF’s expertise is on the importance of movement technique and sport-specific physical requirements of elite sports performance. Moreover, the co-supervision group is composed by researchers/professors from Department of Energy and Process Engineering and Department of Mathematical Science at NTNU.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NAIL Master Thesis pitch 2021/2022] DL Meets Physics: Deep learning to estimate and predict sea current at fish farms in real-time

Project in collaboration with SINTEF Ocean and SINTEF Digital.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[NAIL Master Thesis pitch 2021/2022] Use of GAN and GNN for Unsupervised Anomaly Detection in the Telco domain

The use case and data are provided by Telenor, Norway's leading Telecommunication company. The thesis will be carried out in collaboration with the Analytics and AI team at Telenor Research, a research department inside Telenor which focuses on developing high level models for solving complex Telco use cases and performing state-of-the-art research within AI.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Prediction of Marginal CO2 emission intensity in power grid in Smart Building environment [2021/2022 project in collaboration with Sintef Community]

Problem Description

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

MySQL related projects

There are many opportunities for projects that involve implementation and evaluation in MySQL.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Graph compression and indexing

The project consists of one of several topics related to designing compact/succinct data structures for managing large graphs, and extracting useful information from them efficiently. Typical examples of such huge graphs are genomic sequences and their associated gene expression graphs, web graph and its related SNS data, social networks like Facebook and Twitter, chemical networks, geographical data sets, VLSI graphs, and many more. Traditional data structures do not scale well enough to handle such large graphs, and therefore we need to design new data structures to handle them.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Design of a smart emotion-aware reflection system for teachers

Supervisors: Kshitij Sharma and Sofia Papavlasopoulou

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Design of a smart gaze-aware feedback system for programming

The focus of the thesis is to develop an intelligent feedback system that helps the students while they are programming. This help should be provided in real-time using the eye-tracking data from the student and the log data from the IDE that the student is using. The challenge is to develop a system that is both effective and efficient in helping the students when they are facing difficulties in programming medium-size software.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Design of a smart stress-aware feedback system for programming.

The focus of the thesis is to develop an intelligent feedback system that helps the students while they are programming. This help should be provided in real-time using the eye-tracking data from the student and the log data from the IDE that the student is using. The challenge is to develop a system that is both effective and efficient in helping the students when they are facing difficulties in programming medium-size software.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Using Artificial Intelligence to Predict Student Performance using Eye-tracking Data and Facial Expressions

The focus of the thesis is to develop artificial intelligence pipelines to enhance the prediction of student performance in individual learning tasks, such as video-based learning, programming, assessment. The challenge is to create pipelines that are better suited for a small number of students but with high frequency data (eye-tracking and facial expressions).

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Autorettede diagramoppgaver

I mange emner innen data/informatikk skal studentene blant annet lære å lage modeller, f.eks. ER-modeller, UML-modeller, prosessmodeller, tilstandsdiagrammer. Lignende behov fins også i andre ingeniørdisipliner (f.eks. kretsdiagrammer, konstruksjonsdiagrammer). NTNUs verktøy for digital eksamen (Inspera) har foreløpig begrenset støtte for å kunne tegne diagrammer, som ofte medfører at slike eksamensoppgaver må besvares med blyant og papir og så skannes inn. Dette gjør også at oppgavene nødvendigvis må rettes manuelt, som vil være tidkrevende i emner med mange studenter.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Design av system for team-baserte kodepuslespill

I programmeringsundervisning for nybegynnere er dra-og-slipp oppgaver (i engelsk faglitteratur ofte kalt Parsons Problems) internasjonalt anerkjent som et viktig supplement til vanlige kodeoppgaver. Slike oppgaver er typisk utformet på den måten at alle kodeliner som trengs for løsningen, er gitt - men i omstokket rekkefølge - slik at de må trekkes til riktig posisjon for å få til en løsning. Hvis økt vanskegrad er ønskelig, kan oppgavene også inneholde distraktorer, dvs. kodelinjer som ikke skal brukes i løsningen. For eksempel på Parsons problems, se f.eks. https://www.codio.com/blog/parsons-problems

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Automated indoor tasks assisted by drones

There are many tasks both for indoor and outdoor which lend themselves to being carried out by (semi)autonomous drones. In this project the aim is to develop a system that allows drones to navigate through a building and carry out specific tasks. The system involves a number of main components that allow: dynamically mapping the environment and navigating it without collisions, planning and scheduling missions, managing a fleet of such devices, self-monitor status to ensure returning to the base / charging station.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Automated indoor tasks assisted by rover like robots

There are many tasks both for indoor and outdoor which lend themselves to being carried out by (semi)autonomous robots. In this project the aim is to develop a system that allows rover like robots to navigate through a building and carry out specific tasks. The system involves a number of main components that allow: dynamically mapping the environment and navigating it without collisions, planning and scheduling missions, managing a fleet of such devices, self-monitor status to ensure returning to the base / charging station.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Game of Git (GoG)

The goal of this project is to design and develop an educational game that aims at facilitating learning of core concepts in Git version control system. The student(s) will work to design, implement and evaluate an application (could be web based) that can be used by students learning version control.

Status: Tildelt     Egnet for: En student     Lenke: plink

Improving a mobile app for monitoring group status

The goal of this project is to see how an app can be used to improve group work. This will build on a previous project that already produced such an app.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Knowledge Flow Graphs based Learning Analytics

The goal of this project is to design and develop a web based module for supporting educational activities to be embedded in an existing application (under construction), and that allows creating Knowledge Flow Graphs (KFGs) by annotating lists of Knowledge Components (KCs) and logical dependencies among these in opportune forms. Examples of such KFGs are in https://folk.ntnu.no/damianov/Teaching/TTK4225-2020/hierarchy%20of%20the%20content%20units%20referring%20to%20linear%20algebra.html and in https://folk.ntnu.no/damianov/Teaching/TTK4225-2020/relations%20among%20the%20content%20units.html, and are used for a variety of teaching-related purposes. The application allows, moreover, the creation of quizzes that are annotated with appropriate metadata that allows mapping quizzes to the KFGs, and thus to the KCs that are taught in each topic/subject.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Mobile app for monitoring group status

The goal of this project is to see how an app can be used to improve group work. The sudent(s) will work to design, implement and evaluate an application that can be used by students working in group projects to  easily communicate the health level of their team to the professor.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Web application for generating customized knitted clothing items

The goal of this project is to design and develop a web based user interface for visualizing and accessing a service focused on helping end users to design their own knitting recipes to produce customized clothing items.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Web application for managing educational tasks in Gitlab

The goal of this project is to design and develop a web based user interface for supporting educational activities in courses that use Gitlab for student projects. Typically courses that use Gitlab for managing student groups projects could benefit from having support for setting up and initialize tens or hundreds of repositories, visualize activity, harvest snapshots etc. The application should leverage the Gitlab REST API and potentially Blackboard API to offer the needed features to the staff working in such courses. The student(s) will work to design, implement and evaluate a web application that can be used by professors and teaching assistants. The students should contribute with novel and original ways to deal with the various requirements and especially with the visualization of the data throughout the semester. The student(s) will have to work with the main stakeholders for eliciting requirements.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Web application for visualizing and accessing data from a measurement station

The goal of this project is to design and develop a web based user interface for visualizing and accessing the data from a wastewater measurement station. The station has a number of sensors that constantly produce streams of measurements (weather related such as precipitation and temperature and pipeline related such as water level in the pipe). The sudent(s) will work to design, implement and evaluate a web application that can be used by employees at the station, students and researchers working with this type of data. The students should contribute with novel and original ways to deal with the visualization of the data. The student(s) will have to work with the customer and main stakeholders for eliciting requirements.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Computational Materials

In this project the student(s) will build and test out a prototype of a system of IoT objects that can be programmed without text or visual representations.

https://interactions.acm.org/archive/view/may-june-2017/material-programming

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Cosplay meets Maker Culture

Cosplay is an activity in which participants called cosplayers wear costumes and fashion accessories to represent a specific character. Cosplayers often interact to create a subculture. Favorite sources include anime, cartoons, comic books, manga, television series, and video games.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Cyborg explorations: Adding a human magnetic sense

One way of extending human capabilities is by adding new senses. Products and research exist that add a magnetic sense through a vibrating belt. The belt is connected to a magnetometer and vibrates in the direction of magnetic north. This adds a magnetic sense for the wearer, which improves navigation.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Internet-of-Things exergames

In this project the student(s) will further develop a prototype of an IoT exergame that was developed in the EXACT project.

The project involves both programming, game development and working with physiotherapists in workshops in the UX lab.

 

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Relations we have to digital artefacts

In this research-through-design project, the student(s) will explore the topic of emotional attachment to interactive digital objects.

This will be done through the design of prototypes, user studies and qualitative interviews.

Status: Valgbart     Egnet for: En student     Lenke: plink

A plausible way to understand simple correlations in Norwegian texts

Sticos are developing a bot called @else (http://else.sticos.no) to help Human Resources (HR) departments be more efficient in their work. A lot of their time is consumed by reoccurring questions from their employees and managers. For years HR software have tried to tackle this problem by using a personnel manual. However, the use of it is sparse. The problem lies in accessibility and the fact that is easier to ask the question directly and get a qualified answer to your problem on the fly.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

ChatBots - Dialog interfaces - Text / Phone

We have several systems that make it possible to ask natural language queries over Internet, by SMS or by voice over telephone about various tasks, e.g bus routes or telephone information. You can try yourself by calling +47 7352 1290, or checking http://busstuc.idi.ntnu.no

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Machine learning to identify consistent biological signals

Contemporary biomedical research generates large quantities of data, where each experiment’s design choice contributes to a potential bias in data. More often than not, the analyses deal with the issue by analysing only a single dataset at time, as the biases are difficult to accurately describe. Ironically, combining datasets with different biases can be used to identify and eliminate the biases, and keep only the genuine biological variation.

Status: Valgbart     Egnet for: En student     Lenke: plink

Modeling cell-cycle phase distribution in cell cultures

The cell-cycle is a fundamental molecular process, and disruptions in this process are a hallmark of cancer. When studying eukaryotic cell cycle, the most common approach involves synchronizing a cell culture. In this method, the cells, grown in a medium, are prevented to progress through the cell cycle past a certain phase, usually through use of a chemical agent. The block is then released, and the cells are allowed to progress through the cell cycle.

Status: Valgbart     Egnet for: En student     Lenke: plink

Problems in bioinformatics

The bioinformatics group works on developing and using computational models to predict how changes in gene regulation can control development and cause disease. Towards this end, we develop custom algorithms, statistical simulations, and machine learning-based solutions to analyze and interpret biological data; examples of previous MSc-theses include a genetic programming (GP) approach to predict microRNA target sites, a support vector machine (SVM) approach to identify microRNA genes, and an approach that combines GP and SVMs to identify related proteins.

Status: Valgbart     Egnet for: En student     Lenke: plink

3D Face Reconstruction from 2D Images

3D face reconstruction is the task of transforming 2D image(s) of a face into a 3D model.

Status: Tildelt     Egnet for: En student     Lenke: plink

3D Face Reconstruction from 2D Images

===========Project suggestion for academic year 2021/2022========

Status: Valgbart     Egnet for: En student     Lenke: plink

3D Face Tracking

Face Tracking is a popular research item of the Computer Vision field. Initially, face tracking was based on the extraction of hand-crafted features. The rise of deep learning networks has replaced the hand-crafted features with features automatically extracted by the deep network. Although 2D face tracking has been well studied [1-5], that is not the case for 3D data.
Up until recently, deep networks could not be sufficiently implemented in the case of 3D faces for two reasons: 1) lack of existence of large 3D face datasets and 2) deep networks cannot readily consume 3D data (3D meshes, point clouds etc.) as input.
The 1st issue has been solved with the improvement of 3D data acquisition hardware as well as software capable of producing synthetic 3D facial data [6]. More and more 3D face datasets are thus being produced [7-9].
For resolving the 2nd issue, initially, the methodologies converted the actual 3D data into a 2D modality like depth images or geometry images [10, 11]. Thus, making the data compatible with existing deep networks. This is expected to change in the near future, due to the newest machine learning trend, the so-called Geometric Deep Learning (GDL) [12]. GDL introduces deep networks that can be fed with 3D meshes, point clouds and graphs. Thus, GDL models are very interesting to be studied in order to determine whether they can achieve a performance boost in terms of 3D facial tracking, compared against the 2D case, based on standard evaluation measures.

Status: Tildelt     Egnet for: En student     Lenke: plink

3D Facial Recognition based on Geometric Deep Learning

 

Status: Tildelt     Egnet for: En student     Lenke: plink

3D Facial Recognition based on Geometric Deep Learning

===========Project suggestion for academic year 2021/2022======== 

Status: Valgbart     Egnet for: En student     Lenke: plink

3D-to-2D Facial Recognition based on Geometric Deep Learning

The main goal of this Thesis is a 3D-to-2D Face Recognition approach based on a combination of classical deep learning networks and the new trend in deep learning branch, Geometric Deep Learning Networks.

Status: Valgbart     Egnet for: En student     Lenke: plink

Adventures in Material Point Method (MPM)

Background:
The Material Point Method (MPM) was created in 1994 by Z. Chen et al. as a method for simulating various solids, liquids and gases. In recent times MPM has seen prominent use in graphics simulations, in particular due to Disney's use of the method to simulate snow in Frozen.

Status: Tildelt     Egnet for: En student     Lenke: plink

Animation on the GPU

Animation is an integral part in many 3D applications, especially those who require interactivity. Animation can be thought of as transformations changing over time, giving the illusion of motion for the objects being transformed.

Status: Tildelt     Egnet for: En student     Lenke: plink

Anti-aliasing in Ray Tracing

Background:

Status: Tildelt     Egnet for: En student     Lenke: plink

Continuous 2D Face Authentication System based on webcam input

===========Project suggestion for academic year 2021/2022========

Status: Valgbart     Egnet for: En student     Lenke: plink

Geometric Transformations in WebGL

In computer graphics teaching there is a need to visualize concepts in order to effectively communicate details.
Computer graphics is visual by nature and thus concepts might be easier to grasp if they are visualized.
In addition, an interactive experience would help students to ‘learn by doing’.

Status: Tildelt     Egnet for: En student     Lenke: plink

Incremental 3D Face Reconstruction technique based on webcam facial video

===========Project suggestion for academic year 2021/2022========

Status: Valgbart     Egnet for: En student     Lenke: plink

Integration of a Biometric Access Control System with various levels of security, based on three biometric modalities

===========Project suggestion for academic year 2021/2022========

Status: Valgbart     Egnet for: En student     Lenke: plink

Landscape Modelling Simulation based on User Defined Properties

Landscape modelling is a field with a lot of research done and an increasing demand. The process of designing, modelling and rendering realistic landscapes is a time consuming job and this project aims to improve this process. By automatically generating realistic landscapes and let a designer influence how the terrain is generated could save the designer a lot of time and possibly deliver better results. There are many approaches on how to generate these landscapes and lately there has been a focus on the cooperation between the designer and the software generating the landscape. Some approaches uses real landscape data and machine learning while an other approach is to simulate the world to create the landscape.

Status: Tildelt     Egnet for: En student     Lenke: plink

Ray tracing visualization in VR


The goal of this project is to create an interactive visualization of ray and path tracing that a user can explore in VR. Ray tracing concepts can be hard to learn from descriptions or flat figures, so this project aims at providing an intuitive understanding of ray tracing by letting the user directly “follow” a ray of light on its journey throughout a 3D scene. The application will be made in Unity and will consist of one or more scenes that the user can explore. In addition to ray tracing, general concepts from computer graphics like vector math can also be visualized.

Components:
• Literature review of ray and path tracing
• Writing a path tracer from scratch to gain firsthand knowledge about the intricacies of ray tracing
• Creating the Unity VR visualization of ray tracing

Requirements:
• TDT4230
• Programming with OpenGL and Unity

 

Status: Tildelt     Egnet for: En student     Lenke: plink

Real time volumetric smoke with ray tracing

Volumetric rendering of smoke, clouds and fog is utilized to create photo realistic scenes in computer graphics. However these techniques are offline due to their computational cost.
With modern hardware Ray Tracing becoming closer to mainstream, there appears to be potential for real-time volumetric smoke simulation methods.

Status: Tildelt     Egnet for: En student     Lenke: plink

Real-time Path Tracing

Goal
The main focus of this thesis is to deliver a system for real-time path tracing based on the current
state-of-the-art. More specifically, the following list presents the primary goals:

Status: Tildelt     Egnet for: En student     Lenke: plink

Single view 3D reconstruction for robotic grasping of 3D objects

Robotic grasping is a complex operation that tries to imitate the way humans grasp objects. For robotc grasping to work, it is necessary for the robot to know the shape of the object to be grasped. Humans can ussually infer this shape from a single view, based on previous knowledge and experience. Making robots infer shape from a single view is the aim of this thesis, which includes the following components:

Status: Tildelt     Egnet for: En student     Lenke: plink

Single view 3D reconstruction for robotic manipulation of 3D objects

 

Status: Valgbart     Egnet for: En student     Lenke: plink

Process Mining in Health care

 

This project aims at applying process mining in health care in order to get insight into multiple disease trajectories

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[Learning2Program] Integrasjon av pedagogiske elementer i profesjonelle utviklingsverktøy brukt i utdanningen

På IDI bruker vi i stor grad profesjonelle verktøy i utdanningen, f.eks. VS Code, SceneBuilder og Gitlab, fordi det er viktig å få erfaring med slike, samtidig som de kan bidra til læringen i seg selv. For å forbedre det siste, så kan det tenkes at slike verktøy med fordel kan utvides med elementer som gjør dem bedre egnet pedagogisk sett.

Status: Valgbart     Egnet for: En student     Lenke: plink

[Learning2Program] Learning Analytics for TDT4100

Læringsanalyse er systematisk innsamling og analyse av data om læringsaktiviteter og sammenhengen med læringsutbytte. Målet er å få mer innsikt i hva som skaper god læring og bruke det til å forbedre våre emner.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[Learning2Program] Oppfølging og læringsstøtte i prosjektemner vha. læringsanalyse

Læringsanalyse (learning analytics) er teknikker for å samle inn og analysere data om læringsprosessen, så en kan gi bedre læringsstøtte. Vi jobber med å ta dette i bruk i prosjektemner som IT-prosjektet (IT1901) og programvareutvikling (TDT4140) hvor oppfølging, tilbakemelding og veiledning både er faglig vanskelig og ressurskrevende.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[Play@Campus] Concepts, architecture and technology for location aware and social games

Students are playful and social beings, and why not combine them? Particularly new students need ways of getting to know each other and their new campus, and what better way to do this than to participate in a social, location aware game!?

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[Play@Campus] Prototyping av lokasjonsbevisste spill

Prototyping av lokasjonsbevisste spill er krevende, bl.a. fordi utvikling og evaluering av scenarier er ressurskrevende (mange personer i reelle omgivelser). Oppgaven går ut på å se på enkle og rimelige teknikker for prototyping i mindre skala, som lar seg overføre til reelle omgivelser i stor skale, f.eks. ved innovativ bruk av dukker med RFID-teknologi, gulvstore kart etc.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Computational nanosystems: Matter, Metrics and Models

 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Learning in ,or learning from, self-organizing growing neural network

Keywords: ANN methods and principles, bio-inspired design, evolution, development, genetic algorithm, self-organization, c/c++.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Toward Learning: Exploring search landscape for binarized neural networks

Key words: learning, training, bio-inspired design, ANN, RNN, exploration.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Learning control policies to actuate in Trondheim's air quality

Under the AI4EU project (https://www.ai4eu.eu) a urban simulator of Trondheim is being developed, based on the SUMO traffic simulator. SUMO is an open source traffic simulator that is developed to simulate realistic road networks. As a side effect it also models the pollutant emissions of vehicles which allows us to model the effect of traffic pollution in an urban scenario. Traffic data is publicly available which can be inputted to the traffic simulator and obtain realistic simulated traffic patterns. This simulator serves as a tool that captures the realistic patterns of air quality data and can be used as an environment to train autonomous agents, both in what concerns to improve air quality levels and the quality of the information on pollution levels.
Therefore, the goal of this thesis is the deployment of agents that learn, through reinforcement learning or other planning algorithms, from data coming from this simulator. The goal of the agent can be defined on the setup of the project. Some examples are:

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Optimal placement of air quality sensors

As part of the AI4IoT pilot, a network of air quality low-cost sensors have been deployed throughout the city of Trondheim, with the goal of improving the spatial coverage of air quality information in the city, in comparison to the few industrial sensors previously available.. In this type of scenarios, sensor placement and coverage is crucial to improve the information available in the system and may depend on several factors, including the objectives of the system designer. For instance, the minimal set of sensors to cover a given area, how to place sensors such that the information lost in case of failures is minimal, among others. Therefore, the goal of this thesis is to investigate the sensor placement/coverage problem applied to air quality sensors. Air quality sensors have an added interest in that coverage is not necessarily a well-defined measure (as it is for cameras, for instance) and the preparation of the project will have to investigate how to measure it, as well. Moreover, we have available a urban simulator which models pollution patterns in the city (from traffic and other sources) which can be used as a testing environment for the developed methods during the thesis.

For more information on the topic, feel free to contact Tiago Veiga (tiago.veiga@ntnu.no) or Kerstin Bach (kerstin.bach@ntnu.no).
 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Transfering calibration models between low-cost air quality sensors

As part of the AI4IoT pilot, a network of air quality low-cost sensors have been deployed throughout the city of Trondheim, with the goal of improving the spatial coverage of air quality information in the city, in comparison to the few industrial sensors previously available. While the additional data allows us to gather new information not available before, low-cost sensors are noisier and less reliable. Among others, two of the most important related to that are: 1) how to calibrate a low-cost sensor to provide measurements as good as possible when compared to a reference; 2) how to use low-cost sensors to compute air quality predictions. Calibrations can be based on physical models or data-driven. Whatever the option it must be assessed against a reference sensor, typically a more expensive, industrial sensor. In Trondheim, only a few of these are available and two of the newly deployed low-cost sensors have been co-located on the same location as a reference sensor. At these locations, we can compute calibration/prediction models with the reference data as target. Therefore, the question to be investigated in this thesis is: how to transfer a calibration/prediction model between sensors which don’t have a near reference in the network? Besides air quality data, other related data is publicly available (weather, traffic, etc) and can be taken into account in the project.

For more information on the topic, feel free to contact Tiago Veiga (tiago.veiga@ntnu.no) or Kerstin Bach (kerstin.bach@ntnu.no).
 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[ExerGames] Multi-player pedal-game

The goal of this project is to design and develop new game concepts for a game where an exercise bike is used as a game controller in addition to traditional game input through mulitple buttons. In addition to input from buttons, the player should control the game through using her/his fit moving the pedals. The goal of the game is to both to have fun that can last over time as well as getting a physical exercise. The game should be implemented in Unity using a provided API for the exercise bike controller. 

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

[ExerGames] Multi-player pedal-game

The goal of this project is to design and develop new game concepts for a game where an exercise bike is used as a game controller in addition to traditional game input through mulitple buttons. In addition to input from buttons, the player should control the game through using her/his fit moving the pedals. The goal of the game is to both to have fun that can last over time as well as getting a physical exercise. The game should be implemented in Unity using a provided API for the exercise bike controller. 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[ExerGames] Play to get fit

In this project, the goal is to come up with new game concepts and game technologies for exergames - games where the player carry out physical exercise at the same time. There are several approaches for exergames, and the challenge is to find the ballance between something that is fun to play as well as you get a real physical exercise from playing the game.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

[ExerGames] Play to get fit

In this project, the goal is to come up with new game concepts and game technologies for exergames - games where the player carry out physical exercise at the same time. There are several approaches for exergames, and the challenge is to find the ballance between something that is fun to play as well as you get a real physical exercise from playing the game.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[ExerGames] Play to get fit

In this project, the goal is to come up with new game concepts and game technologies for exergames - games where the player carry out physical exercise at the same time. There are several approaches for exergames, and the challenge is to find the ballance between something that is fun to play as well as you get a real physical exercise from playing the game.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[Game technology] Alternative games

In this project the goal is to prototype an innovative game(s) and test this game(s) on users. The innovation can be in the type of gameplay the game provides, how it combines various game genres, what technology is used to control or play the game, how the social interaction between players are supported, the purpose of the game etc.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

[Game Technology] AR Game to Motivation Socialisation and Physical Activity

In this project, the goal is to develop an game concept that will motivate the users to socialize and being physical activity using Augmented Reality. This project is part of a larger project with the goal of commercialising a concept. 

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

[Game Technology] AR Game to Motivation Socialisation and Physical Activity

In this project, the goal is to develop an game concept that will motivate the users to socialize and being physical activity using Augmented Reality. This project is part of a larger project with the goal of commercialising a concept. 

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

[Lecture Games] Collaborative classroom learning games

The goal of this project is the design, implementation and evaluation of a collaborative learning game, where the students together beat the game and at the same time learn. The game will have to balance engagement and learning, to both make it fun and educational. Another requirement is that the game must be a multiplayer game where all the students in a call can participate at the same time.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Application of machine learning to ILI data denoising

The project is for Master thesis work at AkerSolutions

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Automatic hierarchical sequence segmentation

Given a sequence of nodes and their neighboring similarities, can you segment the nodes automatically?

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Stochastic Multiplicative Updates

This is a project on basic research. We focus on a fundamental optimization problem in machine learning. Successfully solving this problem will improve a wide range of tasks, for example, cluster analysis, topic discovery, signal processing, recommendation.

Status: Tildelt     Egnet for: Gruppe     Lenke: plink

Visualization aided cluster analysis

Cluster analysis is one of the the fundamental tasks in machine learning and data mining. It is used in a wide range of applications such as biology, medicine, world wide web, chemistry, climatology, finance, and social science. In practice, data is often distributed in curved manifolds and the number of clusters is unknown. Conventional clustering methods do not handle such situations. In the project we will attack the problem by using visualization techniques, especially modern nonlinear dimensionality reduction, to facilitate human in the loop of cluster discovery.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Ethical aspects of AI/recommender systems

Smart and personalized systems such as recommender systems (or artificial intelligence in general) keep influencing our daily lives in an increasing rate. In the recent years, researchers became more aware of the ethical challenges in developing such systems in an ethical way such that these systems would treat everyone equal and fair, without any bias or discrimination. Even though these are the topics social scientists have been working on for a long time, defining these concepts as mathematical models, implementing them within AI systems and evaluating the success of these approaches is not an easy task.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Fake News (Disinformation) Detection

The impact of news articles on the society can not be underestimated and as the number of online news are increasing, distinguishing the fake news from real news is becoming a challenge for people. This project focuses on analyzing and tracking news articles from different news sources or social media channels, in order to find an efficient way of detecting fake news.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Noise in the sea: the quest to detect and classify animal and man-made sounds

Akvaplan-niva AS, a daughter company of the Norwegian Institute for Water Research (NIVA). This project is a joint project of Akvaplan-niva AS and NTNU.

Problem description: In recent years, the use of Passive Acoustic Monitoring (PAM) for research and mitigation purposes has dramatically increased, with new studies encouraging its use. Acoustic data provide scientific insights in a broad range of fields including animal vocalizations (biophony) and anthropogenic noise (anthrophony) in the marine environment. Yet, the amount of human effort required to manually identify acoustic features and events rapidly becomes limiting as the size of the data sets increase. Machine learning has the potential to handle large data sets and address many questions in the field of passive acoustic such as automatic detection and classification of sound events. However, the sound from these sources reverberates in the environment which profoundly distorts the original source waveform or can be masked by other sounds overlapping in time and frequencies. Thus the signal recorded usually contains a mixture of highly variable unknown sources, each distorted by the environment in an unknown fashion. Moreover, the noise produced by system events of unmanned autonomous vehicles (gliders) while carrying the hydrophones can saturate the recordings. This issue can mask the detections.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Personalized study assistant for students in higher education

Students in higher education have more freedom to organize their study time and resources. However this situation may cause some difficulties, especially when the students need more guidance and support in certain topics.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Predictive AI for Leid.no

Leid is a technology company that develops and runs a "Rental As A Service Service" targeted at homeowners.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

AI Decisions in Case Processing (“Saksbehandling”) and Unfairness Problem in AI/ML.

“Equals should be treated equally and unequals unequally” - Aristotales

Decision making about peoples’ application for a job, college admission, bank loan, social benefits, etc is increasingly becoming AI-based. At the same time, AI is observed to make ethically wrong decisions. One of the problems in AI ethics is bias which leads to discriminative and unfair decisions.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Evolutionary game theory for Public Good Games – Data sharing

The problem of cooperation has been studied in game theory for long time with a focus on Prisoners dilemma which is a 2-player game. This project topic is about N-player games and focuses on studying and developing mechanism for the emergence of cooperation in societies. There are some known reasons why people “defect” instead of cooperation. There are reasons why they cooperate. This project aims to understand the incentive mechanisms underlying cooperation and to understand their dynamics. This is particulary important for production of public goods.
There are many examples of this overarching problem. Here is one: There is a specific platform for sharing data which may help people in their daily life. This happens only if many people share the data. However, sharing data has a cost (e.g., takes time or may have cognitive load) as well as the benefit from having access to such data. A rational agent would just use the platform for getting the information and let the others do the data-sharing job. The research question is what kind of mechanisms can promote cooperation.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Simulation for Fair Machine Learning - studying long-term effects of fairness policies

It is known that human beings take discriminative decisions, for example,  in hiring, insurance and approval of bank loan on the basis of gender, ethnic origin, race, etc. Computer systems using machine learning (ML) methods are found to repeat the same kind of discrimination - with significant negative impact on peoples' life.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Your project idea - related to AI ethics, in particular, bias, discrimination, unfairness, and diversity.

If you have an idea you want to work with related to bias, (un)fairness, discrimination, etc in Artificial Intelligence systems, email me and we can talk about it.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Informasjonsgjenfinning i semantiske data

Semantisk web data brukes på flere og flere områder og kjennetegnes av å være en lavnivå representasjon av data med eksplisitt semantisk typing. Data lagres typisk i triple-stores og aksesseres med spørringer i sparql. Dette gjør at informasjonen er vanskelig tilgjengelig uten dedikerte applikasjonsspesifikk programvare.

Status: Valgbart     Egnet for: Gruppe     Lenke: plink

Læringsteknologi

Læringsteknologi er programvare og andre teknologiske produkter som understøtter læring og undervisning. Her er det mulighet for selvvalgte oppgaver enten fra studenter eller studenter i samarbeid med fagstab, og prosjekter som kan relateres til enten Excited senter for fremragende utdanning. 

Eksempler på tema for oppgaver:

Status: Valgbart     Egnet for: Gruppe     Lenke: plink
NTNU logo