Autonomous Perception and Deep Learning (DL), inc. game-based data augmentation
Several projects related to this topic are available:
- “Self-driving cars”: Focus on sensor fusion, localization (where is the car), perception (where is everything else, i.e. lane lines, traffic signs and lights, other vehicles, pedestrians etc.) and planning (how to get from a to b). The ambition is to acquire a “mobile dev. lab” like this.
- Autonomous drones: again the ambition is to acquire the best development platform for the perception part during the summer (e.g. DJI dev.). A concrete application is detecting and tracking sheeps in collaboration with Hvasshovd.
- Potentially some perception projects related to the student organizations NTNU Revolve, Ascend NTNU and NTNU Cyborg.
- Data augmentation (generating more data) for training deep neural networks related to autonomous perception using simulations, graphics and gaming.
The projects consist of:
- Studying state-of-the-art solutions within the field
- Development of prototype software
- Testing and validation.
Contact the project provider for more details about the project. The focus of the project will be defined together with the student. It's also possible to work in groups of two.