SENSOR: Imitate Third-Person Expert's Behaviors via Active Sensoring
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866913299339149312 |
|---|---|
| author | Huang, Kaichen Shao, Minghao Wan, Shenghua Sun, Hai-Hang Feng, Shuai Gan, Le Zhan, De-Chuan |
| author_facet | Huang, Kaichen Shao, Minghao Wan, Shenghua Sun, Hai-Hang Feng, Shuai Gan, Le Zhan, De-Chuan |
| contents | In many real-world visual Imitation Learning (IL) scenarios, there is a misalignment between the agent's and the expert's perspectives, which might lead to the failure of imitation. Previous methods have generally solved this problem by domain alignment, which incurs extra computation and storage costs, and these methods fail to handle the \textit{hard cases} where the viewpoint gap is too large. To alleviate the above problems, we introduce active sensoring in the visual IL setting and propose a model-based SENSory imitatOR (SENSOR) to automatically change the agent's perspective to match the expert's. SENSOR jointly learns a world model to capture the dynamics of latent states, a sensor policy to control the camera, and a motor policy to control the agent. Experiments on visual locomotion tasks show that SENSOR can efficiently simulate the expert's perspective and strategy, and outperforms most baseline methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_03386 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | SENSOR: Imitate Third-Person Expert's Behaviors via Active Sensoring Huang, Kaichen Shao, Minghao Wan, Shenghua Sun, Hai-Hang Feng, Shuai Gan, Le Zhan, De-Chuan Robotics Artificial Intelligence Machine Learning In many real-world visual Imitation Learning (IL) scenarios, there is a misalignment between the agent's and the expert's perspectives, which might lead to the failure of imitation. Previous methods have generally solved this problem by domain alignment, which incurs extra computation and storage costs, and these methods fail to handle the \textit{hard cases} where the viewpoint gap is too large. To alleviate the above problems, we introduce active sensoring in the visual IL setting and propose a model-based SENSory imitatOR (SENSOR) to automatically change the agent's perspective to match the expert's. SENSOR jointly learns a world model to capture the dynamics of latent states, a sensor policy to control the camera, and a motor policy to control the agent. Experiments on visual locomotion tasks show that SENSOR can efficiently simulate the expert's perspective and strategy, and outperforms most baseline methods. |
| title | SENSOR: Imitate Third-Person Expert's Behaviors via Active Sensoring |
| topic | Robotics Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2404.03386 |