# Contact-Grounded Policy: Dexterous Visuotactile Policy with Generative Contact Grounding

Source: [Contact-Grounded Policy: Dexterous Visuotactile Policy with Generative Contact Grounding](<https://www.sota2.com/research/paper/69ae53f92d603e6de3e42a7e>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-05-12T05:00:04.826Z

Authors: Zhengtong Xu, Yeping Wang, Ben Abbatematteo, Jom Preechayasomboon, Sonny Chan, Nick Colonnese, Amirhossein H. Memar

Published: 2026-03-05T21:22:49.000Z

## Abstract

Contact-rich dexterous manipulation with multi-finger hands remains an open challenge in robotics because task success depends on multi-point contacts that continuously evolve and are highly sensitive to object geometry, frictional transitions, and slip. Recently, tactile-informed manipulation policies have shown promise. However, most use tactile signals as additional observations rather than modeling contact state or how their action outputs interact with low-level controller dynamics. We present Contact-Grounded Policy (CGP), a visuotactile policy that grounds multi-point contacts by predicting coupled trajectories of actual robot state and tactile feedback, and using a learned contact-consistency mapping to convert these predictions into executable target robot states for a compliance controller. CGP consists of two components: (i) a conditional diffusion model that forecasts future robot state and tactile feedback in a compressed latent space, and (ii) a learned contact-consistency mapping that converts the predicted robot state-tactile pair into executable targets for a compliance controller, enabling it to realize the intended contacts. We evaluate CGP using a physical four-finger Allegro V5 hand with Digit360 fingertip tactile sensors, and a simulated five-finger Tesollo DG-5F hand with dense whole-hand tactile arrays. Across a range of dexterous tasks including in-hand manipulation, delicate grasping, and tool use, CGP outperforms visuomotor and visuotactile diffusion-policy baselines.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Dish Wiping on Sim](<https://www.sota2.com/research/sota/dish-wiping-on-sim.md>): Dish Wiping; SIM; Success Rate: 58.4.

- [Fragile Egg Grasping on Sim](<https://www.sota2.com/research/sota/fragile-egg-grasping-on-sim.md>): Fragile Egg Grasping; SIM; Success Rate: 74.8.

- [In-Hand Box Flipping on Sim](<https://www.sota2.com/research/sota/in-hand-box-flipping-on-sim.md>): In-Hand Box Flipping; SIM; Success Rate: 66.

- [In-Hand Box Flipping on Real](<https://www.sota2.com/research/sota/in-hand-box-flipping-on-real.md>): In-Hand Box Flipping; ReaL; Success Rate: 80.

- [Jar Opening on Real](<https://www.sota2.com/research/sota/jar-opening-on-real.md>): Jar Opening; ReaL; Success Rate: 93.3.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2603.05687>)

[PDF](<https://arxiv.org/pdf/2603.05687.pdf>)