EUVIMEDEUVIMED entdecken
Crossref

Learning Diverse Bimanual Dexterous Manipulation Skills from Human Demonstrations

Bohan Zhou, Haoqi Yuan, Yuhui Fu, Zongqing Lu

Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (34) · 28919-28927

DOI: 10.1609/aaai.v40i34.40127 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Bohan Zhou und 3 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Bimanual dexterous manipulation is a critical yet underexplored area in robotics. Its high-dimensional action space and inherent task complexity present significant challenges for policy learning, and the limited task diversity in existing benchmarks hinders general-purpose skill development. Existing approaches largely depend on reinforcement learning, often constrained by intricately designed reward functions tailored to a narrow set of tasks. In this work, we present a novel approach for efficiently learning diverse bimanual dexterous skills from abundant human demonstrations. Specifically, we introduce BiDexHD, a framework that unifies task construction from existing bimanual datasets and employs teacher-student policy learning to address all tasks. The teacher learns state-based policies using a general two-stage reward function across tasks with shared behaviors, while the student distills the learned multi-task policies into a vision-based policy. With BiDexHD, scalable learning of numerous bimanual dexterous skills from auto-constructed tasks becomes feasible, offering promising advances toward universal bimanual dexterous manipulation. Experiments on TACO tool-using dataset spanning 141 tasks across 6 categories demonstrate a task fulfillment rate of 74.59% on trained tasks and 51.07% on unseen tasks. We further transfer BiDexHD to 11 ARCTIC collaborative tasks and achieve an average of 80.49% task fulfillment rate on trained tasks and 65.99% on unseen task. All empirical results demonstrate the effectiveness and competitive zero-shot generalization capabilities of BiDexHD.

Zitieren

APA

Zhou, B., Yuan, H., Fu, Y., & Lu, Z. (2026). Learning Diverse Bimanual Dexterous Manipulation Skills from Human Demonstrations. Proceedings of the AAAI Conference on Artificial Intelligence, 40(34), 28919-28927. https://doi.org/10.1609/aaai.v40i34.40127

Vancouver

Zhou B, Yuan H, Fu Y, Lu Z. Learning Diverse Bimanual Dexterous Manipulation Skills from Human Demonstrations. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(34):28919-28927. doi:10.1609/aaai.v40i34.40127

BibTeX
@article{Zhou2026,
  author = {Zhou, Bohan and Yuan, Haoqi and Fu, Yuhui and Lu, Zongqing},
  title = {Learning Diverse Bimanual Dexterous Manipulation Skills from Human Demonstrations},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2026},
  volume = {40},
  number = {34},
  pages = {28919--28927},
  doi = {10.1609/aaai.v40i34.40127}
}
Kostenlos starten →