G^3VLA：为视觉-语言-动作模型注入几何归纳偏置

View PDF HTML (experimental)

Abstract:Vision-language-action (VLA) models have rapidly advanced generalist robot manipulation by harnessing the semantic knowledge of pretrained vision-language backbones. However, their visual representations are typically grounded in 2D image coordinates rather than calibrated camera geometry. This limitation is especially pronounced in multi-camera systems, where views are coupled by known intrinsics and extrinsics but processed as independent images. We propose G$^3$VLA, a camera-aware geometric module that injects calibrated structure into the visual-token stream of a pretrained VLA without altering its action space or original action-learning objective. G$^3$VLA combines intrinsic-conditioned ray embeddings, projective positional encoding (PRoPE), and bidirectional cross-view fusion to establish geometrically consistent representations across camera views. Geometric supervision is provided by ground-truth point maps or confidence-gated $\pi^3$X teacher predictions, enabling training without depth sensors or manual geometric annotations. On $\pi_0$, G$^3$VLA consistently improves performance on LIBERO and RoboCasa24 under both supervision regimes, while ground-truth supervision improves five-task RoboTwin2.0 averages with three cameras. Experiments on camera-shifted LIBERO-Goal and real robots demonstrate improved robustness to unseen camera viewpoints. Controlled calibration studies across $\pi_0$ and $\pi_{0.5}$ show that these gains depend on correct camera geometry while remaining stable under mild calibration perturbations. These results demonstrate that grounding VLA visual tokens in calibrated camera geometry provides a simple and general mechanism for improving geometric consistency, multi-view reasoning, and viewpoint robustness in pretrained robot policies.

Submission history

From: Yue Peng [view email]
[v1] Tue, 23 Jun 2026 12:02:36 UTC (7,042 KB)
[v2] Wed, 14 Oct 2026 12:08:55 UTC (4,934 KB)