具身开发者社区

外观

论文arXiv 2608.140282 个月前

AdvDex：通过关节对齐动作和对抗学习从人类演示中学习灵巧操作 ​

AdvDex用统一VLA框架、JAAS与域对抗学习，从人机演示学习灵巧操作，并在手部动作预测和真实操作中验证。

AI 整理
  • 本体 SDK 与控制
  • 数据与采集
  • 具身模型

原文链接：AdvDex: Learning Dexterous Manipulation from Human Demonstrations via Joint-Aligned Actions and Adversarial Learning（arXiv 2608.14028 · 发布日期：2026-08-14）

AdvDex: Learning Dexterous Manipulation from Human Demonstrations via Joint-Aligned Actions and Adversarial Learning

AdvDex用统一VLA框架、JAAS与域对抗学习，从人机演示学习灵巧操作，并在手部动作预测和真实操作中验证。

核心要点 ​

  • 面向灵巧操作规模化难题，缓解机器人演示昂贵、动作空间随具身形态变化及跨具身泛化受限。
  • 提出OmniShare人类操作多模态数据集，提供运动学监督和触觉测量，降低对机器人遥操作的依赖。
  • 设计JAAS规范动作表示，包含SE(3)腕部位姿和15个手指关节，对齐人手、灵巧机器人手和平行夹爪。
  • 结合域对抗学习削弱视觉表征中的具身特定信息，实验显示优于基线并支持零样本人到机器人迁移与少样本适应。

验证范围与局限：验证范围为手部动作预测和真实世界灵巧操作实验，摘要未说明具体任务、平台数量或开源情况。

英文摘要 (Abstract) ​

Dexterous manipulation is a fundamental capability for embodied intelligence, but scaling it remains difficult because robot demonstrations are expensive to collect and action spaces vary across embodiments. Policies trained on heterogeneous data can also entangle task-relevant visual cues with embodiment-specific appearance, limiting cross-embodiment generalization. We present AdvDex, a unified Vision-Language-Action framework for learning dexterous manipulation from human and robot demonstrations. First, we introduce OmniShare, a large-scale multimodal dataset of human manipulation demonstrations that provides high-quality kinematic supervision and tactile measurements while reducing reliance on robot teleoperation. Second, we propose the Joint-Aligned Action Space (JAAS), a canonical action representation comprising an $\mathrm{SE}(3)$ wrist pose and 15 finger joints, thereby functionally aligning human hands, dexterous robot hands, and parallel grippers. Finally, we use domain-adversarial learning to reduce embodiment-specific information in the learned visual representation. Experiments on hand-action prediction and real-world dexterous manipulation show consistent improvements over baselines, effective zero-shot human-to-robot skill transfer, generalization to unseen objects and environments, and data-efficient few-shot adaptation.

作者：Zhiyue Zhao, Jingyi Wu, H X Liu, Mingyu Liu, Liyang Li, Hengdi Zhang, Tong He, Zhengxue Cheng

来源

本页事实对应的公开页面，以原文为准。

  1. arXiv 2608.14028：AdvDex: Learning Dexterous Manipulation from Human Demonstrations via Joint-Aligned Actions and Adversarial Learning访问 2026-10-21