Prompt2Auto: From Motion Prompt to Automated Control via Geometry-Invariant One-Shot Gaussian Process Learning
This addresses the challenge of reducing data requirements and improving generalization for robot skill acquisition from human demonstrations, though it appears incremental as it builds on existing Gaussian process and invariance techniques.
The paper tackles the problem of robots requiring large datasets and failing to generalize across coordinate transformations in learning from demonstration, proposing Prompt2Auto, a geometry-invariant one-shot Gaussian process framework that enables automated control from a single motion prompt, with validation showing it reduces the demonstration burden and generalizes across tasks.
Learning from demonstration allows robots to acquire complex skills from human demonstrations, but conventional approaches often require large datasets and fail to generalize across coordinate transformations. In this paper, we propose Prompt2Auto, a geometry-invariant one-shot Gaussian process (GeoGP) learning framework that enables robots to perform human-guided automated control from a single motion prompt. A dataset-construction strategy based on coordinate transformations is introduced that enforces invariance to translation, rotation, and scaling, while supporting multi-step predictions. Moreover, GeoGP is robust to variations in the user's motion prompt and supports multi-skill autonomy. We validate the proposed approach through numerical simulations with the designed user graphical interface and two real-world robotic experiments, which demonstrate that the proposed method is effective, generalizes across tasks, and significantly reduces the demonstration burden. Project page is available at: https://prompt2auto.github.io