Time Reversal as Self-Supervision
This addresses the problem of reducing supervision in robot learning for manipulation tasks, offering a novel self-supervision approach that is incremental in applying time-reversal to visual control.
The paper tackles the challenge of generalizing robot manipulation to varying conditions by proposing a self-supervised time-reversal model that learns goals and plans without demonstrations, achieving assembly of unseen Tetris-style block pairs using only uncalibrated RGB camera input on a physical robot.
A longstanding challenge in robot learning for manipulation tasks has been the ability to generalize to varying initial conditions, diverse objects, and changing objectives. Learning based approaches have shown promise in producing robust policies, but require heavy supervision to efficiently learn precise control, especially from visual inputs. We propose a novel self-supervision technique that uses time-reversal to learn goals and provide a high level plan to reach them. In particular, we introduce the time-reversal model (TRM), a self-supervised model which explores outward from a set of goal states and learns to predict these trajectories in reverse. This provides a high level plan towards goals, allowing us to learn complex manipulation tasks with no demonstrations or exploration at test time. We test our method on the domain of assembly, specifically the mating of tetris-style block pairs. Using our method operating atop visual model predictive control, we are able to assemble tetris blocks on a physical robot using only uncalibrated RGB camera input, and generalize to unseen block pairs. sites.google.com/view/time-reversal