Model Predictive Adversarial Imitation Learning for Planning from Observation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908859523661824 |
|---|---|
| author | Han, Tyler Bao, Yanda Mehta, Bhaumik Guo, Gabriel Vishwakarma, Anubhav Kang, Emily Jung, Sanghun Scalise, Rosario Zhou, Jason Xu, Bryan Boots, Byron |
| author_facet | Han, Tyler Bao, Yanda Mehta, Bhaumik Guo, Gabriel Vishwakarma, Anubhav Kang, Emily Jung, Sanghun Scalise, Rosario Zhou, Jason Xu, Bryan Boots, Byron |
| contents | Human demonstration data is often ambiguous and incomplete, motivating imitation learning approaches that also exhibit reliable planning behavior. A common paradigm to perform planning-from-demonstration involves learning a reward function via Inverse Reinforcement Learning (IRL) then deploying this reward via Model Predictive Control (MPC). Towards unifying these methods, we derive a replacement of the policy in IRL with a planning-based agent. With connections to Adversarial Imitation Learning, this formulation enables end-to-end interactive learning of planners from observation-only demonstrations. In addition to benefits in interpretability, complexity, and safety, we study and observe significant improvements on sample efficiency, out-of-distribution generalization, and robustness. The study includes evaluations in both simulated control benchmarks and real-world navigation experiments using few-to-single observation-only demonstrations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_21533 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Model Predictive Adversarial Imitation Learning for Planning from Observation Han, Tyler Bao, Yanda Mehta, Bhaumik Guo, Gabriel Vishwakarma, Anubhav Kang, Emily Jung, Sanghun Scalise, Rosario Zhou, Jason Xu, Bryan Boots, Byron Robotics Artificial Intelligence Human demonstration data is often ambiguous and incomplete, motivating imitation learning approaches that also exhibit reliable planning behavior. A common paradigm to perform planning-from-demonstration involves learning a reward function via Inverse Reinforcement Learning (IRL) then deploying this reward via Model Predictive Control (MPC). Towards unifying these methods, we derive a replacement of the policy in IRL with a planning-based agent. With connections to Adversarial Imitation Learning, this formulation enables end-to-end interactive learning of planners from observation-only demonstrations. In addition to benefits in interpretability, complexity, and safety, we study and observe significant improvements on sample efficiency, out-of-distribution generalization, and robustness. The study includes evaluations in both simulated control benchmarks and real-world navigation experiments using few-to-single observation-only demonstrations. |
| title | Model Predictive Adversarial Imitation Learning for Planning from Observation |
| topic | Robotics Artificial Intelligence |
| url | https://arxiv.org/abs/2507.21533 |