Julen Urain De Jesus
Research Interests
- Generative Modeling
- Optimization
- Differential Geometry
- Inverse Reinforcement Learning
Affiliation
TU Darmstadt, Intelligent Autonomous Systems, Computer Science Department
Contact Information
Room E323, Building S2|02, TU Darmstadt, Hochschulstr. 10, 64289 Darmstadt
julen.urain@robot-learning.de
I am currently a postdoctoral researcher at the Intelligent Autonomous Systems lab (IAS) and the German Research Centre for Artificial Intelligence (DFKI). Previously, I interned as a researcher in Nvidia's Seattle Robotics Lab (SRL). I did my Master's degree at UPC and my Master's Thesis at EPFL in the Biorobotics Lab. For my research, I was honoured to be selected as an R:SS Pioneer and finalist for the Georg Girault Ph.D. award for the best European Thesis in Robotics
My research interests lie at the intersection of robotics and machine learning. In particular I explore the combination of fields such as deep generative models, motion planning and control, imitation learning, optimisation, and reinforcement learning. During my PhD, I adapted diffusion models to the Lie group SE(3) to represent 6-DoF grasp pose distributions, explored the composability of energy-based models for reactive motion generation and exploited normalizing flows to learn nonlinear globally stable dynamical systems from demonstrations.
If you are interested in similar topics, I am always looking for collaborations or thesis supervision, so please do not hesitate to contact me.
Software
SE(3)-DiffusionFields: A Python Library to learn diffusion models in the Lie Group SE(3). The repository contains training and sampling algorithms and we provide trained models to generate SE(3) poses to generate grasps for arbitrary objects.
Key References
Cost Learning
-
- Urain, J.; Funk, N.; Peters, J.; Chalvatzaki G (2023). SE(3)-DiffusionFields: Learning smooth cost functions for joint grasp and motion optimization through diffusion, International Conference on Robotics and Automation (ICRA).
-
- Urain, J.; Le, A. T.; Lambert, A.; Chalvatzaki, G.; Boots, B.; Peters, J. (2022). Learning Implicit Priors for Motion Optimization, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).
- Le, A. T.; Urain, J.; Lambert, A.; Chalvatzaki, G.; Boots, B.; Peters, J. (2022). Learning Implicit Priors for Motion Optimization, RSS 2022 Workshop on Implicit Representations for Robotic Manipulation.
Motion Optimization with Energy-Based Models
-
- Urain, J.; Li, A.; Liu, P.; D'Eramo, C.; Peters, J. (2023). Composable energy policies for reactive motion generation and reinforcement learning, International Journal of Robotics Research (IJRR).
- Urain, J.; Li, A.; Liu, P.; D'Eramo, C.; Peters, J. (2021). Composable Energy Policies for Reactive Motion Generation and Reinforcement Learning, Robotics: Science and Systems (RSS).
-
- Urain, J.; Le, A. T.; Lambert, A.; Chalvatzaki, G.; Boots, B.; Peters, J. (2022). Learning Implicit Priors for Motion Optimization, IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).
- Le, A. T.; Urain, J.; Lambert, A.; Chalvatzaki, G.; Boots, B.; Peters, J. (2022). Learning Implicit Priors for Motion Optimization, RSS 2022 Workshop on Implicit Representations for Robotic Manipulation.
Motion Primitives and Dynamic Systems
-
- Urain, J.; Tateo, D.; Peters, J. (2023). Learning Stable Vector Fields on Lie Groups, Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), IEEE R-AL Track.
- Urain, J.; Tateo, D; Peters, J. (2022). Learning Stable Vector Fields on Lie Groups, Robotics and Automation Letters (RA-L).
-
- Urain, J.; Ginesi, M.; Tateo, D.; Peters, J. (2020). ImitationFlow: Learning Deep Stable Stochastic Dynamic Systems by Normalizing Flows, IEEE/RSJ International Conference on Intelligent Robots and Systems.