RoboPianist: Dexterous Piano Playing with Deep Reinforcement Learning
Abstract
Replicating human-like dexterity in robot hands represents one of the largest open problems in robotics. Reinforcement learning is a promising approach that has achieved impressive progress in the last few years; however, the class of problems it has typically addressed corresponds to a rather narrow definition of dexterity as compared to human capabilities. To address this gap, we investigate piano-playing, a skill that challenges even the human limits of dexterity, as a means to test high-dimensional control, and which requires high spatial and temporal precision, and complex finger coordination and planning. We introduce RoboPianist, a system that enables simulated anthropomorphic hands to learn an extensive repertoire of 150 piano pieces where traditional model-based optimization struggles. We additionally introduce an open-sourced environment, benchmark of tasks, interpretable evaluation metrics, and open challenges for future study. Our website featuring videos, code, and datasets is available at https://kzakka.com/robopianist/
Keywords
Cite
@article{arxiv.2304.04150,
title = {RoboPianist: Dexterous Piano Playing with Deep Reinforcement Learning},
author = {Kevin Zakka and Philipp Wu and Laura Smith and Nimrod Gileadi and Taylor Howell and Xue Bin Peng and Sumeet Singh and Yuval Tassa and Pete Florence and Andy Zeng and Pieter Abbeel},
journal= {arXiv preprint arXiv:2304.04150},
year = {2023}
}
Comments
Accepted to the Conference on Robot Learning (CORL) 2023