DexJoCo: A Benchmark and Toolkit for Task-Oriented Dexterous Manipulation on MuJoCo

TL;DR

DexJoCo introduces a comprehensive benchmark and toolkit with 11 tasks for evaluating and advancing task-oriented dexterous manipulation in robotics, emphasizing robustness and generalization.

Contribution

It provides a standardized benchmark with a data collection system, diverse tasks, and extensive empirical analysis to identify challenges in dexterous hand manipulation.

Findings

Modern models struggle with robustness and generalization in dexterous tasks.

The benchmark reveals common limitations of current policies in dexterous manipulation.

Extensive data and analysis highlight key challenges for future research.

Abstract

Achieving human-level manipulation requires dexterous robotic hands capable of complex object interactions. Advancing such capabilities further demands standardized benchmarks for systematic evaluation. However, existing dexterous benchmarks lack tasks that reflect the unique manipulation capabilities of dexterous hands over parallel grippers, as well as comprehensive evaluation pipelines. In this paper, we present DexJoCo, a benchmark and toolkit for task-oriented dexterous manipulation, comprising 11 functionally grounded tasks that evaluate tool-use, bimanual coordination, long-horizon execution, and reasoning. We develop a low-cost data collection system and collect 1.1K trajectories across these tasks, with support for domain randomization to assess robustness. We benchmark modern models under diverse settings, including visual and dynamics randomization, multi-task training, and action-head adaptation. Through extensive empirical analysis, we identify several important insights and common limitations of current policies in dexterous manipulation, highlighting key challenges for future research in dexterous hand robot learning. Project page available at: https://dexjoco.github.io