Translation-Equivariant Self-Supervised Learning for Pitch Estimation with Optimal Transport (2508.01493v1)

Abstract: In this paper, we propose an Optimal Transport objective for learning one-dimensional translation-equivariant systems and demonstrate its applicability to single pitch estimation. Our method provides a theoretically grounded, more numerically stable, and simpler alternative for training state-of-the-art self-supervised pitch estimators.