RoboAtom: Large-scale Atomic-Action Supervision for Robot Learning

arXiv:2610.40207 · cs.RO · Submitted 2026-10-29 · Read on arXiv

cs.RO

Submitted: 2026-10-29

Updated: 2026-10-29

Comments: 31 pages, 12 figures, including references and appendices

License: http://creativecommons.org/publicdomain/zero/1.0/

The gist: Fine-grained action annotations can support skill reuse and compositional generalization in robot policies.

Abstract

Fine-grained action annotations can support skill reuse and compositional generalization in robot policies. However, robot datasets typically provide task-level instructions without subtask annotations. Although some include subtask labels, inconsistent formats and conventions hinder cross-source joint training. We propose RoboAtom-Data, which adds unified, time-aligned atomic-action annotations to existing data from multiple sources, covering 1.45 million trajectories, 6.88 million skill-labeled segments, and 16,882.1 hours. Our pipeline combines a unified protocol, large-scale manual annotation, propagation across repeated executions with the same ordered skill sequence, and multi-stage quality control. We evaluate the dataset through automatic atomic-action annotation and robot control. We construct RoboAtom-Bench to standardize evaluation of skill recognition and temporal boundary localization. Trained on the annotated data, RoboAtom-Anno achieves better automatic annotation performance than all evaluated open- and closed-source models on both RoboAtom-Bench and the external WGO-Bench. For robot control, atomic-action supervision paired with hierarchical execution improves task success over task-level baselines on LIBERO-Long, LIBERO-Plus, and RoboTwin. We will release the data, benchmark, model, and construction pipeline to support further research on skill-level robot learning.

Related papers