Dr. Tim Welschehold

	Junior Research Group Leader
	Postal address:	Albert-Ludwigs-Universität Freiburg Technische Fakultät Robot Learning Lab Georges-Köhler-Allee 080 D-79110 Freiburg i. Br., Germany
	Office:	Buildig 80, 01-002
	Email:	twelsche@cs.uni-freiburg.de
	Phone:	+49 761 203-8005

About me

08/2022 - up to now: Junior Research Group Leader, University of Freiburg, Germany
03/2022 - 07/23: Substitute professor and commissionary head of the Autonomous Intelligent Systems group, University of Freiburg, Germany
03/2020 - 02/2022: Senior Researcher in the Autonomous Intelligent Systems and Robotic Learning labs, University of Freiburg
04/2014 - 02/2019: PhD in Computer Science at the Autonomous Intelligent Systems group of Prof. Dr. Wolfram Burgard, University of Freiburg, Germany
05/2013 - 03/2014: Scientific Researcher, Fraunhofer Institute of Solar Energy Systems
10/2007 - 04/2013: Diploma in Physics with minor in Computer Science, University of Freiburg

Research Interests

Reinforcement learning
Imitation Learning
Mobile Manipulation
Dynamical systems

Current Research Projects

Publications

Google scholar profile

2024

Daniel Honerkamp*, Martin Büchner*, Fabien Despinoy, Tim Welschehold, Abhinav Valada,
Language-Grounded Dynamic Scene Graphs for Interactive Object Search with Mobile Manipulation
IEEE Robotics and Automation Letters (RA-L), 2024.
Download Website BibTeX

@article{honerkamp24momallm,
      title={Language-Grounded Dynamic Scene Graphs for Interactive Object Search with Mobile Manipulation},
      author={Honerkamp, Daniel and Büchner, Martin and Despinoy, Fabien and Welschehold, Tim and Valada, Abhinav},
      journal={arXiv preprint arXiv:2403.08605},
      year={2024}
}

Adrian Röfer, Iman Nematollahi, Tim Welschehold, Wolfram Burgard, Abhinav Valada,
Bayesian Optimization for Sample-Efficient Policy Improvement in Robotic Manipulation
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Abu Dhabi, UAE, 2024
Download Website
Nick Heppert, Max Argus, Tim Welschehold, Thomas Brox, Abhinav Valada,
DITTO: Demonstration Imitation by Trajectory Transformation
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Abu Dhabi, UAE , 2024
arXiv Website

Eugenio Chisari , Nick Heppert , Tim Welschehold, Wolfram Burgard, Abhinav Valada,
CenterGrasp: Object-Aware Implicit Representation Learning for Simultaneous Shape Reconstruction and 6-DoF Grasp Estimation
IEEE Robotics and Automation Letters, vol. 9, no. 6, doi: 10.1109/LRA.2024.3388850., 2024
Download BibTeX

@article{
chisari2023centergrasp,
      title={CenterGrasp: Object-Aware Implicit Representation Learning for Simultaneous Shape Reconstruction and 6-DoF Grasp Estimation}, 
      author={Eugenio Chisari and Nick Heppert and Tim Welschehold and Wolfram Burgard and Abhinav Valada},
      year={2023},
      eprint={2312.08240},
      archivePrefix={arXiv},
      primaryClass={cs.RO}

}

2023

Iman Nematollahi, Kirill Yankov, Wolfram Burgard, Tim Welschehold,
Robot Skill Generalization via Keypoint Integrated Soft Actor-Critic Gaussian Mixture Models
International Symposium on Experimental Robotics (ISER), Chiang Mai, Thailand, 2023.
Download Video Project Page BibTeX

@inproceedings{nematollahi23iser,
    author  = {Iman Nematollahi and Kirill Yankov and Wolfram Burgard and Tim Welschehold},
    title   = {Robot Skill Generalization via Keypoint Integrated Soft Actor-Critic Gaussian Mixture Models},
    booktitle = {Proceedings of the International Symposium on Experimental Robotics (ISER)},
    year    = {2023},
    url={http://ais.informatik.uni-freiburg.de/publications/papers/nematollahi23iser.pdf},
    address = {Chiang Mai, Thailand}
}

Jan Ole von Hartz , Eugenio Chisari , Tim Welschehold, Wolfram Burgard, Joschka Boedecker, Abhinav Valada,
The Treachery of Images: Bayesian Scene Keypoints for Deep Policy Learning in Robotic Manipulation
IEEE Robotics and Automation Letters, vol. 8, no. 11, pp. 6931-6938, doi: 10.1109/LRA.2023.3313917., 2023
Download BibTeX

@article{von_Hartz_2023,
	doi = {10.1109/lra.2023.3313917},  
	url = {https://doi.org/10.1109%2Flra.2023.3313917},  
	year = 2023,
	month = {nov},  
	publisher = {Institute of Electrical and Electronics Engineers ({IEEE})},  
	volume = {8},  
	number = {11},  
	pages = {6931--6938},  
	author = {Jan Ole von Hartz and Eugenio Chisari and Tim Welschehold and Wolfram Burgard and Joschka Boedecker and Abhinav Valada},  
	title = {The Treachery of Images: Bayesian Scene Keypoints for Deep Policy Learning in Robotic Manipulation},  
	journal = {{IEEE} Robotics and Automation Letters}

}

Daniel Honerkamp, Tim Welschehold, Abhinav Valada,
N2M2:Learning Navigation for Arbitrary Mobile Manipulation Motions in Unseen and Dynamic Environments
IEEE Transactions on Robotics (T-RO), 2023.
>> Best Paper Award, IROS 2022 Workshop on Mobile Manipulation and Embodied Intelligence.
Download, Publisher page Video Demo BibTeX

@article{honerkamp22n2m2,
        title={N2M2:Learning Navigation for Arbitrary Mobile Manipulation Motions in Unseen and Dynamic Environments},
         author={Honerkamp, Daniel and Welschehold, Tim and Valada, Abhinav},
         journal={IEEE Transactions on Robotics}, 
         title={N$^{2}$M$^{2}$: Learning Navigation for Arbitrary Mobile Manipulation Motions in Unseen and Dynamic Environments}, 
          year={2023},
          doi={10.1109/TRO.2023.3284346}}

Johan Vertens, Nicolai Dorka, Michael Thompson, Tim Welschehold, Wolfram Burgard,
Improving Deep Dynamics Models for Autonomous Vehicles with Multimodal Latent Mapping of Surfaces
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Detroit, USA, 2023.
Download BibTeX

@article{vertens2023improving,
      title={Improving Deep Dynamics Models for Autonomous Vehicles with Multimodal Latent Mapping of Surfaces}, 
      author={Johan Vertens and Nicolai Dorka and Tim Welschehold and Michael Thompson and Wolfram Burgard},
      journal={IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)}, 
      year={2023},

}

Fabian Schmalstieg, Daniel Honerkamp, Tim Welschehold, Abhinav Valada,
Learning Hierarchical Interactive Multi-Object Search for Mobile Manipulation
IEEE Robotics and Automation Letters (RA-L), 2023, 2023.
Download Website BibTeX

@article{honerkamp23himos,
      title={Learning Hierarchical Interactive Multi-Object Search for Mobile Manipulation},
      author={Schmalstieg, Fabian and Honerkamp, Daniel and Welschehold, Tim and Valada, Abhinav},
      journal={IEEE Robotics and Automation Letters (RA-L), 2023},
      year={2023}
}

Nicolai Dorka, Tim Welschehold, Joschka Boedecker, Wolfram Burgard,
Adaptively Calibrated Critic Estimates for Deep Reinforcement Learning
Robotics and Automation Letters (RA-L), 8(2), 2023. Also at: Workshop on Deep Reinforcement Learning, Conference on Neural Information Processing Systems (NeurIPS) , 2021.
Download BibTeX
```
@article{dorka23ral,
	  title={Adaptively Calibrated Critic Estimates for Deep Reinforcement Learning},
	  author={Dorka, Nicolai and Welschehold, Tim and Boedecker, Joschka and Burgard, Wolfram},
	  journal={Robotics and Automation Letters (RA-L), 8(2)},
	  year={2023}
}
    
```

Abdelrahman Younes*, Daniel Honerkamp*, Tim Welschehold, Abhinav Valada,
Catch Me If You Hear Me: Audio-Visual Navigation in Complex Unmapped Environments with Moving Sounds
Robotics and Automation Letters (RA-L), 8(2), 2023
Download BibTeX
@ARTICLE{10008039, author={Younes, Abdelrahman and Honerkamp, Daniel and Welschehold, Tim and Valada, Abhinav}, journal={IEEE Robotics and Automation Letters}, title={Catch Me if You Hear Me: Audio-Visual Navigation in Complex Unmapped Environments With Moving Sounds}, year={2023}, volume={8}, number={2}, pages={928-935}, doi={10.1109/LRA.2023.3234766}}

Nicolai Dorka, Tim Welschehold, Wolfram Burgard,
Dynamic Update-to-Data Ratio: Minimizing World Model Overfitting
International Conference on Learning Representations (ICLR), 2023. Also at: Decision Awareness in Reinforcement Learning Workshop at the International Conference on Machine Learning (ICML), 2022.
Download BibTeX
```
@article{dorka23iclr,
	  title={Dynamic Update-to-Data Ratio: Minimizing World Model Overfitting},
	  author={Dorka, Nicolai and Welschehold, Tim and Burgard, Wolfram},
	  journal={International Conference on Learning Representations (ICLR)},
	  year={2023}
}
    
```

2022

Fabian Schmalstieg*, Daniel Honerkamp*, Tim Welschehold, Abhinav Valada,
Learning Long-Horizon Robot Exploration Strategies for Multi-Object Search in Continuous Action Spaces
Proceedings of the International Symposium on Robotics Research (ISRR), 2022.
Download BibTeX

@article{fabian22exploration,
	  title={Learning Long-Horizon Robot Exploration Strategies for Multi-Object Search in Continuous Action Spaces},
	  author={Schmalstieg, Fabian and Honerkamp, Daniel and Welschehold, Tim and Valada, Abhinav},
	  journal={Proceedings of the International Symposium on Robotics Research (ISRR)},
	  year={2022}
}
}

Iman Nematollahi, Erick Rosete-Beas, Adrian Röfer, Tim Welschehold, Abhinav Valada, Wolfram Burgard,
Robot Skill Adaptation via Soft Actor-Critic Gaussian Mixture Models
IEEE International Conference on Robotics and Automation (ICRA), Philadelphia, USA, 2022.
Download Video Project Page BibTeX

@article{nematollahi2022icra,
title   = {Robot Skill Adaptation via Soft Actor-Critic Gaussian Mixture Models},
author  = {Iman Nematollahi, Erick Rosete-Beas, Adrian Roefer, Tim Welschehold, Abhinav Valada, Wolfram Burgard},
journal = {IEEE International Conference on Robotics and Automation (ICRA)},
year    = {2022}
}

Jan Ole von Hartz , Eugenio Chisari , Tim Welschehold, Abhinav Valada,
Self-Supervised Learning of Multi-Object Keypoints for Robotic Manipulation
Reinforcement Learning for Contact-Rich Manipulation Workshop at ICRA, Philadelphia, USA, 2022.
Download BibTeX

@article{vHartz22icra_ws,
  url = {https://arxiv.org/pdf/2205.08316.pdf},
  author = {von Hartz, J. O. and Chisari, E. and Welschehold, T. and Valada, A.},
  title = {Self-Supervised Learning of Multi-Object Keypoints for Robotic Manipulation},
  journal = {Reinforcement Learning for Contact-Rich Manipulation Workshop at ICRA},
  year = {2022},
}
}

Eugenio Chisari , Tim Welschehold, Joschka Boedecker, Wolfram Burgard, Abhinav Valada,
Correct Me if I am Wrong: Interactive Learning for Robotic Manipulation
IEEE Robotics and Automation Letters (RA-L), 7(2):3695--3702, 2022.
Download BibTeX

@article{chisari21,
      author={Chisari, Eugenio and Welschehold, Tim and  Boedecker, Joschka and  Burgard, Wolfram and Valada, Abhinav},
      journal={arXiv preprint arXiv:2110.03316},
      title={Correct Me if I am Wrong: Interactive Learning for Robotic Manipulation},
      year={2021},
      volume={},
      number={},
      pages={1-1},
      doi={},
     url={https://arxiv.org/pdf/2110.03316.pdf}
 }

Shengchao Yan , Tim Welschehold, Daniel Büscher, Wolfram Burgard,
Courteous Behavior of Automated Vehicles at Unsignalized Intersections via Reinforcement Learning
IEEE Robotics and Automation Letters (RA-L), 7(1):191--198., 2022.
Download BibTeX

@article{yan21ral,
      author={Yan, Shengchao and Welschehold, Tim and Büscher, Daniel and Burgard, Wolfram},
      journal={IEEE Robotics and Automation Letters},
      title={Courteous Behavior of Automated Vehicles at Unsignalized Intersections via Reinforcement Learning},
      year={2021},
      volume={},
      number={},
      pages={1-1},
      doi={10.1109/LRA.2021.3121807},
     url={http://ais.informatik.uni-freiburg.de/publications/papers/yan21ral.pdf}
 }

Laura Londoño, Adrian Röfer, Tim Welschehold, Abhinav Valada,
Doing Right by Not Doing Wrong in Human-Robot Collaboration
ACM/IEEE International Conference on Human-Robot Interaction Workshop on Fairness and Transparency in Human-Robot Interaction, 2022.
Download BibTeX

@inproceedings{londono22hri,
  title={Doing Right by Not Doing Wrong in Human-Robot Collaboration},
  author={Londoño, Laura and Röfer, Adrian and Welschehold, Tim and Valada, Abhinav},
  booktitle={ACM/IEEE International Conference on Human-Robot Interaction Workshop on Fairness and Transparency in Human-Robot Interaction},
  year={2022}
}

Abdelrahman Younes, Daniel Honerkamp, Tim Welschehold, Abhinav Valada,
SoundSpaces Challenge 2022 - 2nd Place
Conference on Computer Vision and Pattern Recognition (CVPR) Embodied AI Workshop, 2022.
Video Challenge

2021

Daniel Honerkamp, Tim Welschehold, Abhinav Valada,
Learning Kinematic Feasibility for Mobile Manipulation through Deep Reinforcement Learning
in IEEE Robotics and Automation Letters (RA-L) , 2021.
Download Video Demo BibTeX

@article{honerkamp2021learning,
  title={Learning Kinematic Feasibility for Mobile Manipulation through Deep Reinforcement Learning},
  author={Honerkamp, Daniel and Welschehold, Tim and Valada, Abhinav},
  journal={arXiv preprint arXiv:2101.05325},
  year={2021}
}

Abdelrahman Younes, Daniel Honerkamp, Tim Welschehold, Abhinav Valada,
SoundSpaces Challenge - 1st Place
Conference on Computer Vision and Pattern Recognition (CVPR) Embodied AI Workshop, 2021.
Video Challenge

2020

Tim Welschehold
Learning mobile manipulation actions from human demonstrations: an approach to learning and augmenting action models and their integration into task representations
Ph.D. Dissertation
Download

2019

Tim Welschehold, Nichola Abdo, Christian Dornhege, Wolfram Burgard
Combined Task and Action Learning from Human Demonstrations for Mobile Manipulation Applications
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Macau, China, 2019.
Download BibTeX

 @article{twelsche19iros,
  title = {Combined Task and Action Learning from Human Demonstrations for Mobile Manipulation Applications},
  author = {Tim Welschehold and Nichola Abdo and Christian Dornhege and Wolfram Burgard},
  journal = {Proceedings of the International Conference on Intelligent Robots and Systems (IROS)},
  year = {2019},
  address = {Macau, China}
}

Iman Nematollahi , Daniel Kuhner, Tim Welschehold, Wolfram Burgard
Augmenting Action Model Learning by Non-Geometric Features
IEEE International Conference on Robotics and Automation (ICRA), Montreal, Canada, 2019.
Download BibTeX

       @article{nematollahi2019icra,
       title   = {Augmenting Action Model Learning by Non-Geometric Features},
       author  = {Iman Nematollahi, Daniel Kuhner, Tim Welschehold, Wolfram Burgard},
       journal = {IEEE International Conference on Robotics and Automation (ICRA)},
       year    = {2019}
       }

2018

Tim Welschehold, Christian Dornhege, Fabian Paus, Tamim Asfour, Wolfram Burgard
Coupling Mobile Base and End-Effector Motion in Task Space
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Madrid, Spain, 2018.
Download BibTeX

@inproceedings{twelsche18iros,
  title={Coupling mobile base and end-effector motion in task space},
  author={Welschehold, Tim and Dornhege, Christian and Paus, Fabian and Asfour, Tamim and Burgard, Wolfram},
  booktitle=IROS},
  year={2018},
}

Christian Zimmermann, Tim Welschehold, Christian Dornhege, Wolfram Burgard, Thomas Brox
3D Human Pose Estimation in RGBD Images for Robotic Task Learning
IEEE International Conference on Robotics and Automation (ICRA), Brisbane, Australia, 2018.
Download BibTeX

       @InProceedings{zw2018,
       author    = {Christian Zimmermann, Tim Welschehold, Christian Dornhege, Wolfram Burgard and Thomas Brox},
        title     = {3D Human Pose Estimation in RGBD Images for Robotic Task Learning},
        booktitle    = "IEEE International Conference on Robotics and Automation (ICRA)",
        year      = {2018},
        url          = "https://lmb.informatik.uni-freiburg.de/projects/rgbd-pose3d/"
}

2017

Tim Welschehold, Christian Dornhege, Wolfram Burgard
Learning Mobile Manipulation Actions from Human Demonstrations
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Vancouver, Canada, 2017.
Download BibTeX

       @INPROCEEDINGS{8206152,
  author={Welschehold, Tim and Dornhege, Christian and Burgard, Wolfram},
  booktitle={2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)}, 
  title={Learning mobile manipulation actions from human demonstrations}, 
  year={2017},
  pages={3196-3201},
  doi={10.1109/IROS.2017.8206152}}

Tim Welschehold, Nichola Abdo, Christian Dornhege, Wolfram Burgard
Towards Learning Task Intentions from Human Demonstrations
Workshop on “Learning for Collaborative Robotics: Enabling Flexible, Redeployable and Agile Industrial Applications” at the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Vancouver, Canada, 2017.
Download BibTeX

   @article{twelsche17iros_ws,
  title = {Towards Learning Task Intentions from Human Demonstrations},
  author = {Tim Welschehold and Nichola Abdo and Christian Dornhege and Wolfram Burgard},
  journal = {Workshop on “Learning for Collaborative Robotics: Enabling Flexible, Redeployable and Agile Industrial Applications” at the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  year = {2017},
  address = {Vancouver, Canada}
}

2016

Tim Welschehold, Christian Dornhege, Wolfram Burgard
Learning Manipulation Actions from Human Demonstrations
IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Daejeon, South Korea, 2016.
Download BibTeX

    @INPROCEEDINGS{7759555,
  author={Welschehold, Tim and Dornhege, Christian and Burgard, Wolfram},
  booktitle={2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)}, 
  title={Learning manipulation actions from human demonstrations}, 
  year={2016},
  pages={3772-3777},
  doi={10.1109/IROS.2016.7759555}}

Teaching

Seminars

Co-organizer, Deep Learning for Autonomous Systems Seminar, University of Freiburg, WS 2020/21
Co-organizer, Deep Learning Lab, University of Freiburg, SS 2020
Co-organizer, Self-Supervised Learning Seminar, University of Freiburg, SS 2020

Current Students

Model-free RL in mobile manipulation, A. Khalid, Master Project
Hierarchical mobile manipulation, F. Schmalstieg, Master Project

Past Students

Dynamical Audio-Visual Navigation: Catching Moving Unheard Sound Sources in Unmapped 3D Environments, 2021. Master Thesis of Abdelrahman Younes (co-supervised with Daniel Honerkamp)
Deep Reinforcement Learning for Adapting Dynamical Systems, 2021. Master Project of Erick Rosete Beas (co-supervised with Iman Nematollahi )
Augmenting Action Model Learning by Non-Geometric Features, 2018. Master Thesis of Iman Nematollahi (co-supervised with Daniel Kuhner )