Learning Human Utility from Video Demonstrations for Deductive Planning in Robotics
Abstract
We uncouple three components of autonomous behavior (utilitarian value, causal reasoning, and fine motion control) to design an interpretable model of tasks from video demonstrations. Utilitarian value is learned from aggregating human preferences to understand the implicit goal of a task, explaining \textit{why} an action sequence was performed. Causal reasoning is seeded from observations and grows from robot experiences to explain \textit{how} to deductively accomplish sub-goals. And lastly, fine motion control describes \textit{what} actuators to move. In our experiments, a robot learns how to fold t-shirts from visual demonstrations, and proposes a plan (by answering \textit{why}, \textit{how}, and \textit{what}) when folding never-before-seen articles of clothing.
Cite
Text
Shukla et al. "Learning Human Utility from Video Demonstrations for Deductive Planning in Robotics." Proceedings of the 1st Annual Conference on Robot Learning, 2017.Markdown
[Shukla et al. "Learning Human Utility from Video Demonstrations for Deductive Planning in Robotics." Proceedings of the 1st Annual Conference on Robot Learning, 2017.](https://mlanthology.org/corl/2017/shukla2017corl-learning/)BibTeX
@inproceedings{shukla2017corl-learning,
title = {{Learning Human Utility from Video Demonstrations for Deductive Planning in Robotics}},
author = {Shukla, Nishant and He, Yunzhong and Chen, Frank and Zhu, Song-Chun},
booktitle = {Proceedings of the 1st Annual Conference on Robot Learning},
year = {2017},
pages = {448-457},
volume = {78},
url = {https://mlanthology.org/corl/2017/shukla2017corl-learning/}
}