Masterarbeit zu emergenten, breitenadaptiven Repräsentationen mittels räumlich gesteuertem Kanal-Routing (w/m/div.)

Develop and implement a novel Transformer architecture for autonomous driving perception that adaptively adjusts computational depth and width. Design and execute experiments to evaluate model performance and efficiency for tasks like semantic segmentation and object detection, comparing results against baseline models. Analyze emergent network properties regarding resource distribution across image regions and internal feature representation organization. Contribute to cutting-edge research at the intersection of deep learning, computer vision, and autonomous driving to develop safer, scalable perception systems. Requires a Master's degree in Computer Science, Engineering, or Natural Sciences with strong grades, along with deep learning knowledge, PyTorch familiarity, and Python programming skills. On-site presence is mandatory.