One ACT Play: Single Demonstration Behavior Cloning with Action Chunking Transformers
This work addresses the challenge of reducing the number of demonstrations needed for behavior cloning in robotics, which is incremental as it builds on existing methods with specific enhancements.
The paper tackled the problem of learning robot tasks from a single human demonstration by using linear transforms to augment the demonstration and generate trajectories for varied initial conditions, achieving successful completion of three block manipulation tasks with a novel temporal ensembling method that improved robustness.
Learning from human demonstrations (behavior cloning) is a cornerstone of robot learning. However, most behavior cloning algorithms require a large number of demonstrations to learn a task, especially for general tasks that have a large variety of initial conditions. Humans, however, can learn to complete tasks, even complex ones, after only seeing one or two demonstrations. Our work seeks to emulate this ability, using behavior cloning to learn a task given only a single human demonstration. We achieve this goal by using linear transforms to augment the single demonstration, generating a set of trajectories for a wide range of initial conditions. With these demonstrations, we are able to train a behavior cloning agent to successfully complete three block manipulation tasks. Additionally, we developed a novel addition to the temporal ensembling method used by action chunking agents during inference. By incorporating the standard deviation of the action predictions into the ensembling method, our approach is more robust to unforeseen changes in the environment, resulting in significant performance improvements.