Graph-Structured Visual Imitation
This enables robots to quickly learn manipulation skills from visual demonstrations without human annotations or environment instrumentation, though it is incremental in building on existing computer vision advances.
The paper tackles visual imitation for robotic manipulation by framing it as a visual correspondence problem, using factorized representations of entities and spatial arrangements to achieve robust imitation within minutes from a single demonstration, outperforming previous CNN-based methods.
We cast visual imitation as a visual correspondence problem. Our robotic agent is rewarded when its actions result in better matching of relative spatial configurations for corresponding visual entities detected in its workspace and teacher's demonstration. We build upon recent advances in Computer Vision,such as human finger keypoint detectors, object detectors trained on-the-fly with synthetic augmentations, and point detectors supervised by viewpoint changes and learn multiple visual entity detectors for each demonstration without human annotations or robot interactions. We empirically show the proposed factorized visual representations of entities and their spatial arrangements drive successful imitation of a variety of manipulation skills within minutes, using a single demonstration and without any environment instrumentation. It is robust to background clutter and can effectively generalize across environment variations between demonstrator and imitator, greatly outperforming unstructured non-factorized full-frame CNN encodings of previous works.