Back to Reality for Imitation Learning

Edward Johns
Proceedings of the 5th Conference on Robot Learning, PMLR 164:1764-1768, 2022.

Abstract

Imitation learning, and robot learning in general, emerged due to breakthroughs in machine learning, rather than breakthroughs in robotics. As such, evaluation metrics for robot learning are deeply rooted in those for machine learning, and focus primarily on data efficiency. We believe that a better metric for real-world robot learning is time efficiency, which better models the true cost to humans. This is a call to arms to the robot learning community to develop our own evaluation metrics, tailored towards the long-term goals of real-world robotics.

Cite this Paper

BibTeX
@InProceedings{pmlr-v164-johns22a, title = {Back to Reality for Imitation Learning}, author = {Johns, Edward}, booktitle = {Proceedings of the 5th Conference on Robot Learning}, pages = {1764--1768}, 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/johns22a/johns22a.pdf}, url = {https://proceedings.mlr.press/v164/johns22a.html}, abstract = {Imitation learning, and robot learning in general, emerged due to breakthroughs in machine learning, rather than breakthroughs in robotics. As such, evaluation metrics for robot learning are deeply rooted in those for machine learning, and focus primarily on data efficiency. We believe that a better metric for real-world robot learning is time efficiency, which better models the true cost to humans. This is a call to arms to the robot learning community to develop our own evaluation metrics, tailored towards the long-term goals of real-world robotics.} }
Endnote
%0 Conference Paper %T Back to Reality for Imitation Learning %A Edward Johns %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-johns22a %I PMLR %P 1764--1768 %U https://proceedings.mlr.press/v164/johns22a.html %V 164 %X Imitation learning, and robot learning in general, emerged due to breakthroughs in machine learning, rather than breakthroughs in robotics. As such, evaluation metrics for robot learning are deeply rooted in those for machine learning, and focus primarily on data efficiency. We believe that a better metric for real-world robot learning is time efficiency, which better models the true cost to humans. This is a call to arms to the robot learning community to develop our own evaluation metrics, tailored towards the long-term goals of real-world robotics.
APA
Johns, E.. (2022). Back to Reality for Imitation Learning. Proceedings of the 5th Conference on Robot Learning, in Proceedings of Machine Learning Research 164:1764-1768 Available from https://proceedings.mlr.press/v164/johns22a.html.

Related Material