DexCatch: Learning to Catch Arbitrary Objects with Dexterous Hands

Fengbo Lan, Shengjie Wang, Yunzhe Zhang, Haotian Xu, Oluwatosin OluwaPelumi Oseni, Ziye Zhang, Yang Gao, Tao Zhang
Proceedings of The 8th Conference on Robot Learning, PMLR 270:2965-2981, 2025.

Abstract

Achieving human-like dexterous manipulation remains a crucial area of research in robotics. Current research focuses on improving the success rate of pick-and-place tasks. Compared with pick-and-place, throwing-catching behavior has the potential to increase the speed of transporting objects to their destination. However, dynamic dexterous manipulation poses a major challenge for stable control due to a large number of dynamic contacts. In this paper, we propose a Learning-based framework for Throwing-Catching tasks using dexterous hands (LTC). Our method, LTC, achieves a 73% success rate across 45 scenarios (diverse hand poses and objects), and the learned policies demonstrate strong zero-shot transfer performance on unseen objects. Additionally, in tasks where the object in hand faces sideways, an extremely unstable scenario due to the lack of support from the palm, all baselines fail, while our method still achieves a success rate of over 60%.

Cite this Paper

BibTeX
@InProceedings{pmlr-v270-lan25a, title = {DexCatch: Learning to Catch Arbitrary Objects with Dexterous Hands}, author = {Lan, Fengbo and Wang, Shengjie and Zhang, Yunzhe and Xu, Haotian and Oseni, Oluwatosin OluwaPelumi and Zhang, Ziye and Gao, Yang and Zhang, Tao}, booktitle = {Proceedings of The 8th Conference on Robot Learning}, pages = {2965--2981}, year = {2025}, editor = {Agrawal, Pulkit and Kroemer, Oliver and Burgard, Wolfram}, volume = {270}, series = {Proceedings of Machine Learning Research}, month = {06--09 Nov}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v270/main/assets/lan25a/lan25a.pdf}, url = {https://proceedings.mlr.press/v270/lan25a.html}, abstract = {Achieving human-like dexterous manipulation remains a crucial area of research in robotics. Current research focuses on improving the success rate of pick-and-place tasks. Compared with pick-and-place, throwing-catching behavior has the potential to increase the speed of transporting objects to their destination. However, dynamic dexterous manipulation poses a major challenge for stable control due to a large number of dynamic contacts. In this paper, we propose a Learning-based framework for Throwing-Catching tasks using dexterous hands (LTC). Our method, LTC, achieves a 73% success rate across 45 scenarios (diverse hand poses and objects), and the learned policies demonstrate strong zero-shot transfer performance on unseen objects. Additionally, in tasks where the object in hand faces sideways, an extremely unstable scenario due to the lack of support from the palm, all baselines fail, while our method still achieves a success rate of over 60%.} }
Endnote
%0 Conference Paper %T DexCatch: Learning to Catch Arbitrary Objects with Dexterous Hands %A Fengbo Lan %A Shengjie Wang %A Yunzhe Zhang %A Haotian Xu %A Oluwatosin OluwaPelumi Oseni %A Ziye Zhang %A Yang Gao %A Tao Zhang %B Proceedings of The 8th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2025 %E Pulkit Agrawal %E Oliver Kroemer %E Wolfram Burgard %F pmlr-v270-lan25a %I PMLR %P 2965--2981 %U https://proceedings.mlr.press/v270/lan25a.html %V 270 %X Achieving human-like dexterous manipulation remains a crucial area of research in robotics. Current research focuses on improving the success rate of pick-and-place tasks. Compared with pick-and-place, throwing-catching behavior has the potential to increase the speed of transporting objects to their destination. However, dynamic dexterous manipulation poses a major challenge for stable control due to a large number of dynamic contacts. In this paper, we propose a Learning-based framework for Throwing-Catching tasks using dexterous hands (LTC). Our method, LTC, achieves a 73% success rate across 45 scenarios (diverse hand poses and objects), and the learned policies demonstrate strong zero-shot transfer performance on unseen objects. Additionally, in tasks where the object in hand faces sideways, an extremely unstable scenario due to the lack of support from the palm, all baselines fail, while our method still achieves a success rate of over 60%.
APA
Lan, F., Wang, S., Zhang, Y., Xu, H., Oseni, O.O., Zhang, Z., Gao, Y. & Zhang, T.. (2025). DexCatch: Learning to Catch Arbitrary Objects with Dexterous Hands. Proceedings of The 8th Conference on Robot Learning, in Proceedings of Machine Learning Research 270:2965-2981 Available from https://proceedings.mlr.press/v270/lan25a.html.

Related Material