Dex1B: Learning with 1B Demonstrations for Dexterous Manipulation

TL;DR

Dex1B introduces a large-scale, high-quality dataset of one billion diverse demonstrations for dexterous manipulation, generated using a novel generative model that improves feasibility and diversity, validated in simulation and real-world experiments.

Contribution

The paper presents Dex1B, a new dataset of one billion demonstrations created with a generative model that incorporates geometric constraints and diversity conditions.

Findings

Outperforms prior methods on simulation benchmarks

Demonstrates robustness in real-world robot experiments

Provides a scalable dataset for dexterous manipulation tasks

Abstract

Generating large-scale demonstrations for dexterous hand manipulation remains challenging, and several approaches have been proposed in recent years to address this. Among them, generative models have emerged as a promising paradigm, enabling the efficient creation of diverse and physically plausible demonstrations. In this paper, we introduce Dex1B, a large-scale, diverse, and high-quality demonstration dataset produced with generative models. The dataset contains one billion demonstrations for two fundamental tasks: grasping and articulation. To construct it, we propose a generative model that integrates geometric constraints to improve feasibility and applies additional conditions to enhance diversity. We validate the model on both established and newly introduced simulation benchmarks, where it significantly outperforms prior state-of-the-art methods. Furthermore, we demonstrate its effectiveness and robustness through real-world robot experiments. Our project page is at https://jianglongye.com/dex1b