Reinforcement Learning with Time-dependent Goals for Robotic Musicians
This addresses the challenge of temporal goal achievement in robotics for musical performance, though it is incremental as it extends existing goal-conditioned methods to a specific domain.
The paper tackled the problem of robotic musicianship by introducing time-dependent goals in reinforcement learning to handle sequential melodies, and demonstrated that a robotic agent trained in simulation could successfully play the theremin instrument in the real world.
Reinforcement learning is a promising method to accomplish robotic control tasks. The task of playing musical instruments is, however, largely unexplored because it involves the challenge of achieving sequential goals - melodies - that have a temporal dimension. In this paper, we address robotic musicianship by introducing a temporal extension to goal-conditioned reinforcement learning: Time-dependent goals. We demonstrate that these can be used to train a robotic musician to play the theremin instrument. We train the robotic agent in simulation and transfer the acquired policy to a real-world robotic thereminist. Supplemental video: https://youtu.be/jvC9mPzdQN4