外观
论文arXiv 2610.1038421 天前
OpenViTac发布视触觉操作基准与VTLA增强框架
本文由 AI 根据公开内容整理，可能有误，请以原文为准：OpenViTac: Learning and Benchmarking Visuo-Tactile Policies in a Unified Sim-and-Real Framework（arXiv 2610.10384）
核心内容
针对视觉-触觉-语言-动作（VTLA）策略缺乏统一评测标准的现状，研究团队推出OpenViTac视触觉操作基准。该基准构建了仿真与真机配对的测试环境，将接触密集型操作划分为四个触觉相关能力维度，支持对VLA、WAM及VTLA策略进行跨域一致性评估，旨在解决触觉反馈在具身智能中难以量化对比的工程难题。
基于该基准，研究系统探究了不同触觉表征与集成策略对预训练VLA模型性能的影响，并据此提出OpenVTLA触觉增强框架，整合了最优表征与集成方案。同时利用配对设置深入分析Sim-Real协同训练机制，揭示了影响跨域策略学习的关键因子，为视触觉机器人操作提供了从算法验证到模型优化的统一技术底座。
关键看点
- 统一基准：构建仿真-真机配对测试环境，将接触密集操作解耦为四个触觉能力维度，支持VLA/WAM/VTLA策略跨域一致性评测
- 增强框架：提出OpenVTLA框架，通过对比实验筛选最优触觉表征与集成策略，显著提升预训练VLA模型在复杂操作中的性能表现
- 协同训练：利用配对数据研究Sim-Real Co-training机制，量化分析仿真与真实数据混合比例及域差异对跨域策略迁移效率的具体影响
- 评测对象：覆盖VLA、WAM及VTLA三类主流具身策略架构，提供标准化接口与数据集，解决触觉感知算法缺乏横向对比基准的行业痛点
工程与产业观察
该工作确立了视触觉操作“仿真-真机配对”评测范式，解决了触觉算法难以跨平台对齐的难题。OpenVTLA框架验证了触觉表征选择对VLA模型性能的决定性作用，为具身大模型接入触觉传感提供了工程化路径。其Sim-Real协同训练分析结论，可直接指导低成本仿真数据与高价值真机数据的配比策略，加速触觉灵巧操作落地。
英文摘要 (Abstract)
Tactile feedback provides embodied agents with physical information beyond visual observations, enabling more reliable interaction with the real world. However, despite the rapid progress of vision-tactile-language-action (VTLA) policies, there remains a lack of unified benchmarks for evaluating tactile-enabled robot manipulation across simulation and the real world. To address this gap, we introduce OpenViTac, a visuo-tactile manipulation benchmark for evaluating robot policies across simulation and the real world. OpenViTac organizes contact-rich manipulation into four tactile-relevant capability dimensions and provides paired simulation-real-world settings for consistent evaluation of VLA, WAM, and VTLA policies. Building upon this benchmark, we investigate how different tactile representations and integration strategies affect the performance of pretrained VLA models. Correspondingly, we introduce OpenVTLA, a tactile augmentation framework that combines the best-performing representation and integration strategy. Furthermore, we leverage the paired benchmark setting to study sim-real co-training and analyze factors affecting cross-domain policy learning. Together, OpenViTac provides a unified platform for evaluating and advancing visuo-tactile robot manipulation.
作者：Yifan Wu, Qin Li, Nan Min, Guojin Zhong, Haoyu Zhao, Zhiyuan Li, Houze Xu, Shengqi Xu, Xingyao Lin, Zijie Diao, Zhaoxiang Liu, Shiguo Lian, Shunlin Lu, Shihao Zhao, Ziyi Ye, Zuxuan Wu, Yu-Gang Jiang
来源
本页事实对应的公开页面，以原文为准。