Zhao, S;
Xu, Y;
Kasaei, M;
Khadem, M;
Li, Z;
(2024)
Neural ODE-based Imitation Learning (NODE-IL): Data-Efficient Imitation Learning for Long-Horizon Multi-Skill Robot Manipulation.
In:
IEEE International Conference on Intelligent Robots and Systems.
(pp. pp. 8524-8530).
IEEE: Abu Dhabi, United Arab Emirates.
Preview |
PDF
Neural ODE-based Imitation Learning (NODE-IL) - Data-Efficient Imitation Learning for Long-Horizon Multi-Skill Robot Manipulation.pdf - Accepted Version Download (2MB) | Preview |
Abstract
In robotics, acquiring new skills through Imitation Learning (IL) is crucial for handling diverse complex tasks. However, model-free IL faces challenges of data inefficiency and prolonged training time, whereas model-based methods struggle to obtain accurate nonlinear models. To address these challenges, we developed Neural ODE-based Imitation Learning (NODE-IL), a novel model-based imitation learning framework that employs Neural Ordinary Differential Equations (Neural ODEs) for learning task dynamics and control policies. NODE-IL comprises (1) Dynamic-NODE for learning the continuous differentiable task's transition dynamics model, and (2) Control-NODE for learning a long-horizon control policy in an MPC fashion, which are trained holistically. Extensively evaluated on challenging manipulation tasks, NODE-IL demonstrates significant advantages in data efficiency, requiring less than 70 samples to achieve robust performance. It outperforms Behavioral Cloning from Observation (BCO) and Gaussian Process Imitation Learning (GP-IL) methods, achieving 70% higher average success rate, and reducing translation errors for high-precision tasks, which demonstrates its robustness and accuracy, as an effective and efficient imitation learning approach for learning complex manipulation tasks.
| Type: | Proceedings paper |
|---|---|
| Title: | Neural ODE-based Imitation Learning (NODE-IL): Data-Efficient Imitation Learning for Long-Horizon Multi-Skill Robot Manipulation |
| Event: | 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) |
| Dates: | 14 Oct 2024 - 18 Oct 2024 |
| Open access status: | An open access version is available from UCL Discovery |
| DOI: | 10.1109/IROS58592.2024.10802736 |
| Publisher version: | https://doi.org/10.1109/iros58592.2024.10802736 |
| Language: | English |
| Additional information: | This version is the author accepted manuscript. For information on re-use, please refer to the publisher’s terms and conditions. |
| Keywords: | Training , Robust control , Translation , Accuracy , Imitation learning , Scalability , Mathematical models , Data models , Robustness , Manipulator dynamics |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Computer Science |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10205033 |
Loading...
Loading...Loading...Loading...Archive Staff Only
 |
View Item |
