視覚運動制御のための対比的行動-画像事前学習
Contrastive Action-Image Pre-training for Visuomotor Control
大規模な人間の第一人称ビデオから手のキーポイントを行動の代理として抽出し、対比学習でロボットの視覚エンコーダを事前学習する手法を提案。実機の器用操作タスクで既存手法を30%以上上回る性能を達成した。
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: Yuvan Sharma, Dantong Niu, Anirudh Pai, Zekai Wang, Zhuoyang Liu, Baifeng Shi, Stefano Saravalle, Boning Shao, Ruijie Zheng, Jing Wang, Konstantinos Kallidromitis, Yusuke Kato, Fabio Galasso, Yuke Zhu, Danfei Xu, Linxi "Jim" Fan, Jitendra Malik, Trevor Darrell, Roei Herzig
分類: cs.RO, cs.CV
原文アブストラクト
Existing vision encoders for robotics face a fundamental bottleneck: robotic datasets lack the scale necessary for large-scale pre-training. Prior work circumvents this data scarcity by turning to internet-scale image and language data or egocentric human video. While these models show promise, neither paradigm learns from paired vision and action data, which downstream visuomotor control policies require. However, robot trajectories, the most direct source of this paired signal, are not available at pre-training scale, motivating us to extract action signals from abundant human video instead. To this end, we introduce CAIP (Contrastive Action-Image Pre-training), a vision encoder that treats human hand poses from large-scale egocentric video as a proxy for end-effector actions. By extracting 3D hand keypoints, a representation that aligns naturally with downstream robot action spaces, CAIP learns a unified action-image representation through a contrastive objective. Leveraging 32,041 hours of egocentric human video and only 88 hours of robotic manipulation data, CAIP outperforms state-of-the-art vision encoders including DINOv2, SigLIP, MVP, and R3M. Evaluated on a challenging real-world dexterous manipulation setup using Dexmate Vega and Sharpa Wave hands, CAIP yields performance gains of more than 30% on tasks involving folding, pouring, and fine-grained manipulation. Our results show that our method of contrastive action-centric pre-training yields a scalable path to achieving robust visual representations better suited for physical interaction.
