TeViR: Text-to-Video Reward with Diffusion Models for Efficient Reinforcement Learning
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866915357152772096 |
|---|---|
| author | Chen, Yuhui Li, Haoran Jiang, Zhennan Wen, Haowei Zhao, Dongbin |
| author_facet | Chen, Yuhui Li, Haoran Jiang, Zhennan Wen, Haowei Zhao, Dongbin |
| contents | Developing scalable and generalizable reward engineering for reinforcement learning (RL) is crucial for creating general-purpose agents, especially in the challenging domain of robotic manipulation. While recent advances in reward engineering with Vision-Language Models (VLMs) have shown promise, their sparse reward nature significantly limits sample efficiency. This paper introduces TeViR, a novel method that leverages a pre-trained text-to-video diffusion model to generate dense rewards by comparing the predicted image sequence with current observations. Experimental results across 11 complex robotic tasks demonstrate that TeViR outperforms traditional methods leveraging sparse rewards and other state-of-the-art (SOTA) methods, achieving better sample efficiency and performance without ground truth environmental rewards. TeViR's ability to efficiently guide agents in complex environments highlights its potential to advance reinforcement learning applications in robotic manipulation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_19769 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | TeViR: Text-to-Video Reward with Diffusion Models for Efficient Reinforcement Learning Chen, Yuhui Li, Haoran Jiang, Zhennan Wen, Haowei Zhao, Dongbin Robotics Artificial Intelligence Developing scalable and generalizable reward engineering for reinforcement learning (RL) is crucial for creating general-purpose agents, especially in the challenging domain of robotic manipulation. While recent advances in reward engineering with Vision-Language Models (VLMs) have shown promise, their sparse reward nature significantly limits sample efficiency. This paper introduces TeViR, a novel method that leverages a pre-trained text-to-video diffusion model to generate dense rewards by comparing the predicted image sequence with current observations. Experimental results across 11 complex robotic tasks demonstrate that TeViR outperforms traditional methods leveraging sparse rewards and other state-of-the-art (SOTA) methods, achieving better sample efficiency and performance without ground truth environmental rewards. TeViR's ability to efficiently guide agents in complex environments highlights its potential to advance reinforcement learning applications in robotic manipulation. |
| title | TeViR: Text-to-Video Reward with Diffusion Models for Efficient Reinforcement Learning |
| topic | Robotics Artificial Intelligence |
| url | https://arxiv.org/abs/2505.19769 |