関係的空間アライメントによる視覚-言語-行動モデルの強化
Relational Spatial Alignment for Vision-Language-Action Models
VLAモデルに3D空間認識を注入する際、点ごとの一致ではなくペア関係のアライメントを行うことで表現のランク崩壊を防ぎ、操作タスクの空間理解を向上させた研究。
詳しい要約
1. どんなもの？
2. 先行研究と比べてどこがすごい？
3. 技術・手法の肝は？
4. どうやって有効だと検証した？
5. 議論はある？
6. 次に読むべき論文は？
※ AIが要旨から生成した要約です。正確性は原文をご確認ください。
著者: Chaoran Zhu, Hengyi Wang, Chenglin Cui, Yazhe Wan, Taein Kwon, Changjae Oh
分類: cs.RO
原文アブストラクト
Spatial understanding is essential for robots to perceive their surrounding environments and execute actions. Existing approaches typically encode 3D awareness into vision-language-action (VLA) models by directly aligning their representations with pre-trained 3D foundation models. However, we observe that such direct geometric feature alignment essentially exploits 3D foundation priors as a form of low-rank regularization in the feature space, resulting in severe rank collapse of the pre-trained VLA representations. In contrast, we shift the focus from enforcing point-wise equality to aligning pair-wise relationships. We propose a representation-alignment method that enables the injection of 3D spatial awareness without compromising the model's original expressivity. Extensive experiments on public benchmarks, our new spatial reasoning benchmark, and real-world manipulation scenarios demonstrate that our approach effectively incorporates 3D spatial representations into VLA reasoning, achieving robust performance and enhanced spatial awareness across diverse manipulation tasks.
