面向视觉-语言-动作模型的跨手部潜表征
Cross-Hand Latent Representation for Vision-Language-Action Models
把不同灵巧手的关节动作映射到共享的具身无关潜动作空间，可直接插入标准 VLA 架构，为视觉研究者提供一种跨具身数据复用与动作表征对齐的即插即用方案。
原文依据 a vision-language-action framework integrated with a unified latent action space shared across diverse dexterous hands
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论灵巧操作是真实世界机器人自主性的关键，正如人手协调在人类日常活动中处于核心地位。人类依赖丰富的多模态感知——视觉、声音以及语言引导的意图——来完成灵巧动作，这促使机器人的操作走向基于视觉、以语言为条件的系统。然而，为灵巧操作训练可靠的视觉-语言-动作（VLA）模型，需要在多种机器人手上采集大规模示范数据。此外，随着新的灵巧具身形态快速涌现，为每一种形态单独采集数据既昂贵又不切实际，因此产生了对可扩展跨具身学习的需求。我们提出 \ourmethod，一个视觉-语言-动作框架，集成了在多种灵巧手之间共享的统一潜动作空间。这一具身无关的潜空间可直接插入标准 VLA 架构，从而实现无缝的跨具身训练，并高效复用已有数据与新采集的数据。实验结果表明，\ourmethod 持续优于在原始关节空间中运行的基线 VLA 模型，确立了其作为可扩展跨具身灵巧操作有效方案的地位。
英文原文摘要 1224 字符
Dexterous manipulation is essential for real-world robot autonomy, mirroring the central role of human hand coordination in daily activity. Humans rely on rich multimodal perception--vision, sound, and language-guided intent--to perform dexterous actions, motivating vision-based, language-conditioned manipulation systems for robots. However, training reliable vision-language-action (VLA) models for dexterous manipulation requires large-scale demonstrations across many robotic hands. In addition, as new dexterous embodiments appear rapidly, collecting data for each becomes costly and impractical, creating a need for scalable cross-embodiment learning. We introduce \ourmethod, a vision-language-action framework integrated with a unified latent action space shared across diverse dexterous hands. This embodiment-invariant latent space is directly pluggable into standard VLA architectures, enabling seamless cross-embodiment training and efficient reuse of both existing and newly collected data. Experimental results demonstrate that \ourmethod consistently outperforms baseline VLA models operating in raw joint spaces, establishing it as an effective solution for scalable cross-embodiment dexterous manipulation.