Skill Acquisition via Automated Multi-Coordinate Cost Balancing
This addresses skill acquisition from demonstrations for robotics or motion planning, but appears incremental as it builds on existing coordinate-based optimization approaches.
The paper tackles the problem of acquiring point-to-point movement skills from demonstrations by proposing Multi-Coordinate Cost Balancing (MCCB), which encodes demonstrations in multiple differential coordinates and solves a convex optimization problem with learned weighting factors. The method demonstrates effectiveness on one handwriting dataset and three complex skill datasets.
We propose a learning framework, named Multi-Coordinate Cost Balancing (MCCB), to address the problem of acquiring point-to-point movement skills from demonstrations. MCCB encodes demonstrations simultaneously in multiple differential coordinates that specify local geometric properties. MCCB generates reproductions by solving a convex optimization problem with a multi-coordinate cost function and linear constraints on the reproductions, such as initial, target, and via points. Further, since the relative importance of each coordinate system in the cost function might be unknown for a given skill, MCCB learns optimal weighting factors that balance the cost function. We demonstrate the effectiveness of MCCB via detailed experiments conducted on one handwriting dataset and three complex skill datasets.