TacCoRL: 通过仿真将触觉反馈集成到视觉-语言-动作模型中
TacCoRL: Integrating Tactile Feedback into VLA via Simulation
AI总结：
提出TacCoRL框架，通过仿真与真实联合训练和强化学习，将触觉反馈注入视觉-语言-动作策略，在接触密集型任务中平均成功率提升22.5%。
摘要：
视觉-语言-动作（VLA）模型为机器人操作提供了强大的视觉、语言和动作先验，但仅凭视觉观察往往缺失接触密集型任务所需的局部接触状态。我们提出TacCoRL，一个可扩展的框架，将触觉反馈注入VLA策略，并通过仿真-真实联合训练和基于仿真的强化学习（RL）进行改进，无需大规模触觉预训练或广泛的真实世界接触探索。关键思想不仅是添加触觉作为输入，而是学习在接近失败状态下接触读数应如何调节动作响应，这些状态在演示中罕见且在硬件上收集风险高。我们使用真实对齐的仿真器作为接触交互的闭环训练环境。混合的仿真和真实轨迹首先在预训练策略中热启动触觉条件动作。具有可验证任务奖励的强化学习随后通过仿真接触回滚优化策略。它强化导致任务完成的触觉条件动作，而真实轨迹上的监督目标将精炼策略锚定到部署的视觉、触觉和动作分布。所得策略直接转移到真实机器人，无需特权仿真状态或在线真实世界RL。在四个双臂接触密集型任务中，最终的视觉-触觉策略平均成功率达到72.5%，而基线为50.0%。结果视频和更多细节见此链接。
英文摘要：
Vision-language-action (VLA) models provide strong visual, language, and action priors for robot manipulation, but visual observations alone often miss the local contact state required for contact-rich tasks. We present TacCoRL, a scalable framework that injects Tactile feedback into VLA policies and improves them through sim-real Co-training and simulation-based reinforcement learning (RL), without requiring large-scale tactile pretraining or extensive real-world contact exploration. The key idea is not only adding touch as an input, but learning how contact readings should modulate action responses in near-failure states that are rare in demonstrations and risky to collect on hardware. We use a real-aligned simulator as a closed-loop training environment for contact interaction. Mixed simulated and real trajectories first warm-start tactile-conditioned actions in the pretrained policy. Reinforcement learning with verifiable task rewards then optimizes the policy using simulated contact rollouts. It reinforces tactile-conditioned actions that lead to task completion, while a supervised objective on real trajectories keeps the refined policy anchored to deployment visual, tactile, and action distributions. The resulting policy transfers directly to the real robot without privileged simulation state or online real-world RL. Across four bimanual contact-rich tasks, the final visuo-tactile policy achieves an average success rate of 72.5%, compared to baseline of 50.0%. Result videos and more details are available at https://tac-corl.github.io/
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
