Automated Curriculum Design for High-dimensional Human Motor Learning

TL;DR

This paper introduces an automated, model-based curriculum design framework that accelerates high-dimensional human motor learning, validated through simulations and human studies with significant efficiency gains.

Contribution

It presents a novel framework combining human motor learning models and real-time skill estimation with control algorithms for personalized curriculum design.

Findings

Accelerates skill acquisition by approximately 23% compared to random curricula.

Achieves about 17% improvement over heuristic-based curricula.

Validated effectiveness through simulations and human-subject experiments with a hand exoskeleton.

Abstract

Designing effective practice schedules for high-dimensional motor learning tasks remains a challenge, especially when skill states are unobservable and task performance may not reflect the true learning. We propose an automated curriculum design framework that combines a human motor learning model and personalized real-time skill estimation with Stochastic Nonlinear Model Predictive Control in \emph{de-novo} (novel) motor learning paradigms. We validated our framework both through simulations and human-subject studies (N = 36) using a hand exoskeleton. Our proposed approach accelerates skill acquisition by $\sim23\%$, and ${\sim17\%}$ when compared to a random curriculum and a performance heuristics-based curriculum, respectively. These significant gains in learning efficiency highlight the potential of model-based, individualized curricula for motor rehabilitation and complex skill training.