StereoVLA: Enhancing Vision-Language-Action Models with Stereo Vision
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_ | 1866915694766981120 |
|---|---|
| author | Deng, Shengliang Yan, Mi Zheng, Yixin Su, Jiayi Zhang, Wenhao Zhao, Xiaoguang Cui, Heming Zhang, Zhizheng Wang, He |
| author_facet | Deng, Shengliang Yan, Mi Zheng, Yixin Su, Jiayi Zhang, Wenhao Zhao, Xiaoguang Cui, Heming Zhang, Zhizheng Wang, He |
| contents | Stereo cameras closely mimic human binocular vision, providing rich spatial cues critical for precise robotic manipulation. Despite their advantage, the adoption of stereo vision in vision-language-action models (VLAs) remains underexplored. In this work, we present StereoVLA, a VLA model that leverages rich geometric cues from stereo vision. We propose a novel Geometric-Semantic Feature Extraction module that utilizes vision foundation models to extract and fuse two key features: 1) geometric features from subtle stereo-view differences for spatial perception; 2) semantic-rich features from the monocular view for instruction following. Additionally, we propose an auxiliary Interaction-Region Depth Estimation task to further enhance spatial perception and accelerate model convergence. Extensive experiments show that our approach outperforms baselines by a large margin in diverse tasks under the stereo setting and demonstrates strong robustness to camera pose variations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_21970 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | StereoVLA: Enhancing Vision-Language-Action Models with Stereo Vision Deng, Shengliang Yan, Mi Zheng, Yixin Su, Jiayi Zhang, Wenhao Zhao, Xiaoguang Cui, Heming Zhang, Zhizheng Wang, He Robotics Stereo cameras closely mimic human binocular vision, providing rich spatial cues critical for precise robotic manipulation. Despite their advantage, the adoption of stereo vision in vision-language-action models (VLAs) remains underexplored. In this work, we present StereoVLA, a VLA model that leverages rich geometric cues from stereo vision. We propose a novel Geometric-Semantic Feature Extraction module that utilizes vision foundation models to extract and fuse two key features: 1) geometric features from subtle stereo-view differences for spatial perception; 2) semantic-rich features from the monocular view for instruction following. Additionally, we propose an auxiliary Interaction-Region Depth Estimation task to further enhance spatial perception and accelerate model convergence. Extensive experiments show that our approach outperforms baselines by a large margin in diverse tasks under the stereo setting and demonstrates strong robustness to camera pose variations. |
| title | StereoVLA: Enhancing Vision-Language-Action Models with Stereo Vision |
| topic | Robotics |
| url | https://arxiv.org/abs/2512.21970 |