Training Dynamics-Aware Multi-Factor Curriculum Learning for Target Speaker Extraction

TL;DR

This paper introduces a multi-factor curriculum learning approach for target speaker extraction that adapts to training dynamics, improving performance especially in challenging multi-speaker situations.

Contribution

It proposes a joint curriculum learning strategy combined with a visualization framework to guide data sampling based on observed training dynamics.

Findings

Enhanced extraction accuracy in complex scenarios

Identification of data regions with different learning difficulties

Improved training efficiency through dynamic curriculum adjustment

Abstract

Target speaker extraction (TSE) aims to isolate a specific speaker's voice from multi-speaker mixtures. Despite strong benchmark results, real-world performance often degrades due to different interacting factors. Previous curriculum learning approaches for TSE typically address these factors separately, failing to capture their complex interactions and relying on predefined difficulty factors that may not align with actual model learning behavior. To address this challenge, we first propose a multi-factor curriculum learning strategy that jointly schedules SNR thresholds, speaker counts, overlap ratios, and synthetic/real proportions, enabling progressive learning from simple to complex scenarios. However, determining optimal scheduling without predefined assumptions remains challenging. We therefore introduce TSE-Datamap, a visualization framework that grounds curriculum design in observed training dynamics by tracking confidence and variability across training epochs. Our analysis reveals three characteristic data regions: (i) easy-to-learn examples where models consistently perform well, (ii) ambiguous examples where models oscillate between alternative predictions, and (iii) hard-to-learn examples where models persistently struggle. Guided by these data-driven insights, our methods improve extraction results over random sampling, with particularly strong gains in challenging multi-speaker scenarios.