RIKEN Center for Advanced Intelligence Project Robot Learning Team

Team Leader: Takayuki Osa (Ph.D.)

Japanese Page

Research Summary

The Robot Learning Team works on algorithms and methodologies for robotic agents to learn skills for performing tasks in physical environments. Achieving complex and adaptive behaviors in robotic systems requires active and structured data collection, learning theories and algorithms in high-dimensional state-action spaces, and robust systems against disturbances in real-world environments. Our team aims to establish a foundation for building robotic systems that function in the real world.

Main Research Fields

Informatics

Related Research Fields

Engineering
Humanities & Social Sciences
Interdisciplinary Science & Engineering
Mathematical & Physical Sciences
Intelligent robotics
Intelligent informatics
Robotics and intelligent system

Keywords

Robot Learning
Reinforcement Learning
Imitation Learning
Motion Planning

Selected Publications

Papers with an asterisk(*) are based on research conducted outside of RIKEN.

1. T. Osa and T. Harada
"Discovering Multiple Solutions from a Single Task in Offline Reinforcement Learning"
Proceedings of the International Conference on Machine Learning (ICML), 2024.
2. J. Ackermann, T. Osa, and M. Sugiyama
"Offline Reinforcement Learning from Datasets with Structured Non-Stationarity"
Proceedings of the Reinforcement Learning Conference (RLC), 2024.
3. M. Omura, T. Osa, Y. Mukuta, T. Harada
"Stabilizing Extreme Q-learning by Maclaurin Expansion"
Proceedings of the Reinforcement Learning Conference (RLC), 2024.
4. *T. Osa and T. Harada
"Robustifying a Policy in Multi-Agent RL with Diverse Cooperative Behavior and Adversarial Style Sampling for Assistive Tasks"
Proceedings of the IEEE International Conferences on Robotics and Automation (ICRA), 2024.
5. N. Morihira, P. Deo, M. Bhadu, A. Hayashi, T. Hasegawa, S. Otsubo, T. Osa
"Touch-Based Manipulation with Multi-Fingered Robot using Off-policy RL and Temporal Contrastive Learning"
Proceedings of the IEEE International Conferences on Robotics and Automation (ICRA), 2024.
6. T. Osa
"Motion Planning by Learning the Solution Manifold in Trajectory Optimization"
The International Journal of Robotics Research, Vol. 41, No. 3, pp. 291-311, 2022.
7. T. Osa
"Multimodal Trajectory Optimization for Motion Planning"
The International Journal of Robotics Research, Vol. 39 No. 8, pp. 983–-1001, 2020.
8. *T. Osa, V. Tangkaratt, M. Sugiyama.
"Hierarchical Reinforcement Learning via Advantage-Weighted Information Maximization"
Proceedings of the International Conference on Learning Representations (ICLR), 2019.
9. *T. Osa, J. Pajarinen, G. Neumann, J. A. Bagnell, P Abbeel, and J. Peters.
"An Algorithmic Perspective on Imitation Learning"
Trends and Foundations in Robotics, Vol. 7: No. 1-2, pp 1-179, 2018.
10. *T. Osa, N. Sugita, and M. Mitsuishi.
"Online Trajectory Planning in Dynamic Environments for Surgical Task Automation"
Proceedings of Robotics: Science and Systems (R:SS), 2014.

Lab Members

Principal investigator

Takayuki Osa: Team Leader

Contact Information

Nihonbashi 1-chome Mitsui Building, 15th floor,
1-4-1 Nihonbashi,
Chuo-ku, Tokyo
103-0027, Japan
Email: takayuki.osa@riken.jp