Stefan Weigand

Stefan Werner is a research associate in Robotics and Control at the Safe Autonomous Systems Chair of TU Dortmund. After receiving his B.S. and M.S. in Computer Science from Paderborn University, he worked as a Machine Learning Engineer on embedded vision algorithms at SICK AG. He is currently pursuing a doctorate on safe sim-to-real transfer of reinforcement learning methods and focuses on topics at the intersection of computer vision, robot simulation, and embodied intelligence.

Current research topics:

• Learning transferable control policies for real robots in simulation despite sim-to-real gaps
• Generalization of reinforcement learning to unseen visual domains
• Guaranteeing safe and robust control under domain shifts with reinforcement learning

Special interests:

• Fusing complementary sensor modalities to mitigate limitations of single sensors so that the robustness of learned control policies improves.