Raphael Trumpp


Address  

Chair of Cyber-Physical Systems in Production Engineering
School of Engineering and Design, Technical University of Munich
MW2102a, Boltzmannstraße 15, 85748 Garching b. München

Phone  

+49 89 289 55183

Email  

raphael.trumpp@tum.de

Profile  

Main Research

  • Residual policy learning and deep reinforcement learning
  • Methods for neural network pruning
  • Sim2real gap for learning-based methods
  • Applications: Autonomous driving (CARLA) and racing (F1TENTH)

Links  

        

Publications


Journals

Mirco Theile, Daniele Bernardini, Raphael Trumpp, Cristina Piazza, Marco Caccamo, Alberto L. Sangiovanni-Vincentelli,  "Learning to Generate All Feasible Actions",  in IEEE Access,  Mar 2024.

Conferences

Mirco Theile and Lukas Dirnberger and Raphael Trumpp and Marco Caccamo and Alberto Sangiovanni-Vincentelli,  "Action Mapping for Reinforcement Learning in Continuous Environments with Constraints",  in Reinforcement Learning Journal (To appear.),  Aug 2025.

Raphael Trumpp, Ansgar Schäfftlein, Mirco Theile, Marco Caccamo,  "Impoola: The Power of Average Pooling for Image-based Deep Reinforcement Learning",  in Reinforcement Learning Journal,  Aug 2025 (To appear.).

Raphael Trumpp, Ehsan Javanmardi, Jin Nakazato, Manabu Tsukada, Marco Caccamo,  "RaceMOP: Mapless online path planning for multi-agent autonomous racing using residual policy learning",  in IEEE/RSJ International Conference on Intelligent Robots and Systems 2024,  Abu Dhabi, UAE,  Oct 2024.

Raphael Trumpp, Martin Büchner, Abhinav Valada, Marco Caccamo,  "Efficient Learning of Urban Driving Policies Using Bird's-Eye-View State Representations",  in IEEE International Conference on Intelligent Transportation Systems 2023,  Bilbao, Spain,  Sep 2023.

Raphael Trumpp, Denis Hoornaert, Marco Caccamo,  "Residual Policy Learning for Vehicle Control of Autonomous Racing Cars",  in IEEE Intelligent Vehicles Symposium 2023,  Anchorage, USA,  Jun 2023.

Raphael Trumpp, Harald Bayerlein, David Gesbert,  "Modeling Interactions of Autonomous Vehicles and Pedestrians with Deep Multi-Agent Reinforcement Learning for Collision Avoidance",  in IEEE Intelligent Vehicles Symposium 2022,  Aachen, Germany,  Jun 2022.

Preprints

Benjamin David Evans, Raphael Trumpp, Marco Caccamo, Felix Jahncke, Johannes Betz, Hendrik Willem Jordaan, Herman Arnold Engelbrecht,  "Unifying f1tenth autonomous racing: Survey, methods and benchmarks",  in arXiv,  .