Augmented Feedback Training for Overcoming the Learning Plateau of Motor Expertise

Abstract

Skilled performers often encounter a plateau in which further practice yields little improvement of intricate sensorimotor skills. Overcoming such limits requires novel training paradigms that can engage performers in novel ways of exploring and refining their actions. Here, we introduce a training pipeline that integrates high-precision motion sensing with augmented feedback to enhance expert-level motor learning. Using piano performance as a model, trained pianists practiced imitating a prize-winning expert's performance of a technically demanding movement sequence. While conventional auditory-based learning offered limited benefit, augmenting practice with trial-by-trial visualizations of discrepancies between the pianist's own movements and those of the expert enabled learners to refine their performance. This training induced richer movement exploration, facilitated closer convergence toward expert motion patterns, and produced perceptible improvements in sound quality evaluated by expert pianists. These findings demonstrate that augmented feedback can break performance plateaus in experts by providing externally-sensed, task-specific information that expands exploration beyond habitual training strategies. Such augmented learning promises new applications across domains where expertise is constrained by entrenched motor habits, including musical performance, athletics, and surgical training.