Co-GAIL: Learning Diverse Strategies for Human-Robot Collaboration

Chen Wang, Claudia Pérez-D’Arpino, Danfei Xu, Li Fei-Fei, Karen Liu, Silvio Savarese
Proceedings of the 5th Conference on Robot Learning, PMLR 164:1279-1290, 2022.

Abstract

We present a method for learning human-robot collaboration policy from human-human collaboration demonstrations. An effective robot assistant must learn to handle diverse human behaviors shown in the demonstrations and be robust when the humans adjust their strategies during online task execution. Our method co-optimizes a human policy and a robot policy in an interactive learning process: the human policy learns to generate diverse and plausible collaborative behaviors from demonstrations while the robot policy learns to assist by estimating the unobserved latent strategy of its human collaborator. Across a 2D strategy game, a human-robot handover task, and a multi-step collaborative manipulation task, our method outperforms the alternatives in both simulated evaluations and when executing the tasks with a real human operator in-the-loop. Supplementary materials and videos at https://sites.google.com/view/cogail/home

Cite this Paper

BibTeX
@InProceedings{pmlr-v164-wang22h, title = {Co-GAIL: Learning Diverse Strategies for Human-Robot Collaboration}, author = {Wang, Chen and P\'erez-D'Arpino, Claudia and Xu, Danfei and Fei-Fei, Li and Liu, Karen and Savarese, Silvio}, booktitle = {Proceedings of the 5th Conference on Robot Learning}, pages = {1279--1290}, 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/wang22h/wang22h.pdf}, url = {https://proceedings.mlr.press/v164/wang22h.html}, abstract = {We present a method for learning human-robot collaboration policy from human-human collaboration demonstrations. An effective robot assistant must learn to handle diverse human behaviors shown in the demonstrations and be robust when the humans adjust their strategies during online task execution. Our method co-optimizes a human policy and a robot policy in an interactive learning process: the human policy learns to generate diverse and plausible collaborative behaviors from demonstrations while the robot policy learns to assist by estimating the unobserved latent strategy of its human collaborator. Across a 2D strategy game, a human-robot handover task, and a multi-step collaborative manipulation task, our method outperforms the alternatives in both simulated evaluations and when executing the tasks with a real human operator in-the-loop. Supplementary materials and videos at https://sites.google.com/view/cogail/home} }
Endnote
%0 Conference Paper %T Co-GAIL: Learning Diverse Strategies for Human-Robot Collaboration %A Chen Wang %A Claudia Pérez-D’Arpino %A Danfei Xu %A Li Fei-Fei %A Karen Liu %A Silvio Savarese %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-wang22h %I PMLR %P 1279--1290 %U https://proceedings.mlr.press/v164/wang22h.html %V 164 %X We present a method for learning human-robot collaboration policy from human-human collaboration demonstrations. An effective robot assistant must learn to handle diverse human behaviors shown in the demonstrations and be robust when the humans adjust their strategies during online task execution. Our method co-optimizes a human policy and a robot policy in an interactive learning process: the human policy learns to generate diverse and plausible collaborative behaviors from demonstrations while the robot policy learns to assist by estimating the unobserved latent strategy of its human collaborator. Across a 2D strategy game, a human-robot handover task, and a multi-step collaborative manipulation task, our method outperforms the alternatives in both simulated evaluations and when executing the tasks with a real human operator in-the-loop. Supplementary materials and videos at https://sites.google.com/view/cogail/home
APA
Wang, C., Pérez-D’Arpino, C., Xu, D., Fei-Fei, L., Liu, K. & Savarese, S.. (2022). Co-GAIL: Learning Diverse Strategies for Human-Robot Collaboration. Proceedings of the 5th Conference on Robot Learning, in Proceedings of Machine Learning Research 164:1279-1290 Available from https://proceedings.mlr.press/v164/wang22h.html.

Related Material