arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.10698cs.CVcs.AI
收藏

AugVLA-3D:基于深度的特征增强用于视觉-语言-动作模型

AugVLA-3D: Depth-Driven Feature Augmentation for Vision-Language-Action Models

Guangdong Laboratory of Artificial Intelligence and Digital Economy (SZ)(广东省人工智能与数字经济实验室(深圳));Department of Computer Science and Engineering, Southern University of Science and Technology(南方科技大学计算机科学与工程系);School of Communication and Information Engineering, Shanghai University(上海大学通信与信息工程学院);School of Information Technology, Carleton University(卡尔顿大学信息技术学院)

Zhifeng Rao, Wenlong Chen, Lei Xie, Xia Hua, Dongfu Yin, Zhen Tian, F. Richard Yu

AI总结:

本文提出AugVLA-3D框架,通过整合深度估计提升视觉-语言-动作模型的3D特征表示,增强模型在复杂3D环境中的感知与动作预测能力。

摘要:

视觉-语言-动作(VLA)模型近期在机器人感知与控制中取得显著进展,但现有方法主要依赖于使用2D图像训练的视觉语言模型,限制了其在复杂3D环境中的空间理解与动作定位能力。为解决这一限制,我们提出了一种新的框架,将深度估计整合到VLA模型中以丰富3D特征表示。具体而言,我们采用名为VGGT的深度估计基线,从标准RGB输入中提取几何感知的3D线索,从而高效利用现有大规模2D数据集,同时隐式恢复3D结构信息。为进一步增强这些深度衍生特征的可靠性,我们引入了一个名为动作助手的新模块,通过动作先验约束学习的3D表示,并确保其与下游控制任务的一致性。通过将增强的3D特征与传统2D视觉标记融合,我们的方法显著提高了VLA模型的泛化能力和鲁棒性。实验结果表明,所提方法不仅增强了在几何模糊场景中的感知能力,还实现了更优的动作预测准确性。本工作突显了基于深度的数据增强和辅助专家监督在弥合2D观察与3D感知决策之间差距方面的潜力。

英文摘要:

Vision-Language-Action (VLA) models have recently achieved remarkable progress in robotic perception and control, yet most existing approaches primarily rely on VLM trained using 2D images, which limits their spatial understanding and action grounding in complex 3D environments. To address this limitation, we propose a novel framework that integrates depth estimation into VLA models to enrich 3D feature representations. Specifically, we employ a depth estimation baseline called VGGT to extract geometry-aware 3D cues from standard RGB inputs, enabling efficient utilization of existing large-scale 2D datasets while implicitly recovering 3D structural information. To further enhance the reliability of these depth-derived features, we introduce a new module called action assistant, which constrains the learned 3D representations with action priors and ensures their consistency with downstream control tasks. By fusing the enhanced 3D features with conventional 2D visual tokens, our approach significantly improves the generalization ability and robustness of VLA models. Experimental results demonstrate that the proposed method not only strengthens perception in geometrically ambiguous scenarios but also leads to superior action prediction accuracy. This work highlights the potential of depth-driven data augmentation and auxiliary expert supervision for bridging the gap between 2D observations and 3D-aware decision-making in robotic systems.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑