Contact Coverage-Guided Exploration for General-Purpose Dexterous Manipulation

2026-03-11

AI Summary

This work addresses the lack of a general-purpose reward mechanism in dexterous manipulation, where existing approaches often rely on task-specific priors. The authors propose Contact Coverage-Guided Exploration (CCGE), which introduces contact coverage as a universal exploration signal for the first time. By modeling contact states between fingers and object regions, CCGE combines count-based coverage rewards with energy-based reachability rewards to drive efficient exploration of diverse contact patterns. A learned hash encoding is employed to discretize object states, enabling effective exploration without requiring task-specific prior knowledge. Experiments demonstrate that CCGE significantly improves both training efficiency and success rates across a range of dexterous manipulation tasks, and the learned policies successfully transfer to real-world robotic systems.

Technology Category

• Intelligent Robots: Manipulation
• Search and Optimization: Sampling/Simulation-based Search
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Deep Reinforcement learning (DRL) has achieved remarkable success in domains with well-defined reward structures, such as Atari games and locomotion. In contrast, dexterous manipulation lacks general-purpose reward formulations and typically depends on task-specific, handcrafted priors to guide hand-object interactions. We propose Contact Coverage-Guided Exploration (CCGE), a general exploration method designed for general-purpose dexterous manipulation tasks. CCGE represents contact state as the intersection between object surface points and predefined hand keypoints, encouraging dexterous hands to discover diverse and novel contact patterns, namely which fingers contact which object regions. It maintains a contact counter conditioned on discretized object states obtained via learned hash codes, capturing how frequently each finger interacts with different object regions. This counter is leveraged in two complementary ways: (1) to assign a count-based contact coverage reward that promotes exploration of novel contact patterns, and (2) an energy-based reaching reward that guides the agent toward under-explored contact regions. We evaluate CCGE on a diverse set of dexterous manipulation tasks, including cluttered object singulation, constrained object retrieval, in-hand reorientation, and bimanual manipulation. Experimental results show that CCGE substantially improves training efficiency and success rates over existing exploration methods, and that the contact patterns learned with CCGE transfer robustly to real-world robotic systems. Project page is https://contact-coverage-guided-exploration.github.io.

Problem

• dexterous manipulation
• reward formulation
• contact patterns
• exploration
• general-purpose

Innovation

• Contact Coverage
• Dexterous Manipulation
• Exploration Reward
• Hash-based State Discretization
• Deep Reinforcement Learning