FOCAL-VLA：サブタスク誘導型幾何蒸留と暗黙的ワールドモデリングによる視覚言語行動モデル
FOCAL-VLA: Subtask-Guided Geometry Distillation and Implicit World Modeling for Vision-Language-Action Models
サブタスクに関連する領域に絞って幾何知識を蒸留し、将来の3D変化を暗黙的にモデル化することで、精密かつ長期的なロボット操作を可能にするVLAフレームワークを提案。
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: Zhiyuan Gao, Di Wen, Yanxiang Zhan, Mohammad Khoshnazar, Jeroen Schäfer, Kunyu Peng, Michael Beetz
分類: cs.RO, cs.AI, cs.CV
原文アブストラクト
Vision-language-action (VLA) models built on pretrained vision-language models have demonstrated strong performance across diverse robotic manipulation tasks. However, VLA models that directly map current 2D observations to actions often lack sufficient spatial and temporal understanding, limiting their performance in precise and long-horizon manipulation. Recent methods enhance VLA models through geometric supervision and future-state prediction across the entire scene. However, these methods can suffer from redundant scene information, distracting the model from learning the geometry and dynamics relevant to the current interaction. To address this issue, we propose FOCAL-VLA, a framework that combines subtask-guided geometry distillation with implicit world modeling to learn representations of current spatial structure and future interaction dynamics. To focus geometric learning on the current subtask, we transfer geometric knowledge from VGGT to the VLA model by aligning geometry latents with features from subtask-relevant image regions. To capture the future 3D evolution of the current interaction, we incorporate implicit world modeling using Track4World features from current and future demonstration frames. The two complementary representations jointly guide action generation without running VGGT or Track4World at inference time. Experiments show that FOCAL-VLA outperforms baselines on both simulation benchmarks and real-world manipulation tasks. Project website: https://zhiyuan-gao.github.io/FOCAL-VLA/.