AI RESEARCH
Refinement of Accelerated Demonstrations via Incremental Iterative Reference Learning Control for Fast Contact-Rich Imitation Learning
arXiv CS.AI
•
ArXi:2604.16850v1 Announce Type: cross Fast execution of contact-rich manipulation is critical for practical deployment, yet providing fast nstrations for imitation learning (IL) remains challenging: humans cannot nstrate at high speed, and naively accelerating nstrations alters contact dynamics and induces large tracking errors. We present a method to autonomously refine time-accelerated nstrations by repurposing Iterative Reference Learning Control (IRLC) to iteratively update the reference trajectory from observed tracking errors.