-
To Style or Not to Style? A Comparison Between Data Augmentation and Style Supression for Domain Generalization
Important note: this project is provisional, and the details may be subject to change.
Description: During this project, you will use an existing network for semantic segmentation and study whether its generalization capacity can be improved by performing either data augmentation or style supression.
notset
-
Towards Modeling Vehicle Interactions for the purpose of Path Prediction
Important note: this project is provisional, and the details may be subject to change.
Description: During this project, you will be able to model the interactions between vehicles to improve trajectory prediction of those vehicles.
notset
-
Teach a driving agent what it cannot see
Important note: this project is provisional, and the details may be subject to change.
Description: During this project, you will implement, train and evaluate a Reinforcement Learning method for driving in a 3D simulation enviornment
notset
-
Pedestrian Path Prediction via Imitation Learning with Inverse Reinforcement Learning
Important note: this project is provisional, and the details may be subject to change.
Description: During this project, you will apply a state-of-the-art imitation learning algorithm to the task of pedestrian path prediction.
notset
-
Accuracy and Efficiency Improvements for Fast Panoptic Segmentation
Important note: this project is not available anymore. It remains visible to show the type of student projects available at our group.
Description: During this project, you will improve an existing network for panoptic segmentation, to make it more accurate and efficient.
notset
