A System for General In-Hand Object Re-Orientation

Tao Chen, Jie Xu, Pulkit Agrawal
Proceedings of the 5th Conference on Robot Learning, PMLR 164:297-307, 2022.

Abstract

In-hand object reorientation has been a challenging problem in robotics due to high dimensional actuation space and the frequent change in contact state between the fingers and the objects. We present a simple model-free framework that can learn to reorient objects with both the hand facing upwards and downwards. We demonstrate the capability of reorienting over 2000 geometrically different objects in both cases. The learned policies show strong zero-shot transfer performance on new objects. We provide evidence that these policies are amenable to real-world operation by distilling them to use observations easily available in the real world. The videos of the learned policies are available at: https://taochenshh.github.io/projects/in-hand-reorientation.

Cite this Paper


BibTeX
@InProceedings{pmlr-v164-chen22a, title = {A System for General In-Hand Object Re-Orientation}, author = {Chen, Tao and Xu, Jie and Agrawal, Pulkit}, booktitle = {Proceedings of the 5th Conference on Robot Learning}, pages = {297--307}, year = {2022}, editor = {Faust, Aleksandra and Hsu, David and Neumann, Gerhard}, volume = {164}, series = {Proceedings of Machine Learning Research}, month = {08--11 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v164/chen22a/chen22a.pdf}, url = {https://proceedings.mlr.press/v164/chen22a.html}, abstract = {In-hand object reorientation has been a challenging problem in robotics due to high dimensional actuation space and the frequent change in contact state between the fingers and the objects. We present a simple model-free framework that can learn to reorient objects with both the hand facing upwards and downwards. We demonstrate the capability of reorienting over $2000$ geometrically different objects in both cases. The learned policies show strong zero-shot transfer performance on new objects. We provide evidence that these policies are amenable to real-world operation by distilling them to use observations easily available in the real world. The videos of the learned policies are available at: https://taochenshh.github.io/projects/in-hand-reorientation.} }
Endnote
%0 Conference Paper %T A System for General In-Hand Object Re-Orientation %A Tao Chen %A Jie Xu %A Pulkit Agrawal %B Proceedings of the 5th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2022 %E Aleksandra Faust %E David Hsu %E Gerhard Neumann %F pmlr-v164-chen22a %I PMLR %P 297--307 %U https://proceedings.mlr.press/v164/chen22a.html %V 164 %X In-hand object reorientation has been a challenging problem in robotics due to high dimensional actuation space and the frequent change in contact state between the fingers and the objects. We present a simple model-free framework that can learn to reorient objects with both the hand facing upwards and downwards. We demonstrate the capability of reorienting over $2000$ geometrically different objects in both cases. The learned policies show strong zero-shot transfer performance on new objects. We provide evidence that these policies are amenable to real-world operation by distilling them to use observations easily available in the real world. The videos of the learned policies are available at: https://taochenshh.github.io/projects/in-hand-reorientation.
APA
Chen, T., Xu, J. & Agrawal, P.. (2022). A System for General In-Hand Object Re-Orientation. Proceedings of the 5th Conference on Robot Learning, in Proceedings of Machine Learning Research 164:297-307 Available from https://proceedings.mlr.press/v164/chen22a.html.

Related Material