TouchGuide: Inference-Time Steering of Visuomotor Policies via Touch Guidance
For robotic manipulation tasks requiring fine-grained contact, TouchGuide provides a method to incorporate tactile feedback without retraining the base policy, enabling better performance in contact-rich scenarios.
TouchGuide introduces a cross-policy visuo-tactile fusion paradigm that steers pre-trained visuomotor policies at inference time using tactile guidance from a Contact Physical Model, achieving significant improvements over state-of-the-art on five contact-rich tasks (e.g., shoe lacing, chip handover).
Fine-grained and contact-rich manipulation remain challenging for robots, largely due to the underutilization of tactile feedback. To address this, we introduce TouchGuide, a novel cross-policy visuo-tactile fusion paradigm that fuses modalities within a low-dimensional action space. Specifically, TouchGuide operates in two stages to guide a pre-trained diffusion or flow-matching visuomotor policy at inference time. First, the policy produces a coarse, visually-plausible action using only visual inputs during early sampling. Second, a task-specific Contact Physical Model (CPM) provides tactile guidance to steer and refine the action, ensuring it aligns with realistic physical contact conditions. Trained through contrastive learning on limited expert demonstrations, the CPM provides a tactile-informed feasibility score to steer the sampling process toward refined actions that satisfy physical contact constraints. Furthermore, to facilitate TouchGuide training with high-quality and cost-effective data, we introduce TacUMI, a data collection system. TacUMI achieves a favorable trade-off between precision and affordability; by leveraging rigid fingertips, it obtains direct tactile feedback, thereby enabling the collection of reliable tactile data. Extensive experiments on five challenging contact-rich tasks, such as shoe lacing and chip handover, show that TouchGuide consistently and significantly outperforms state-of-the-art visuo-tactile policies.