Contrastive Predictive Coding Based Feature for Automatic Speaker Verification

TL;DR

This work explores the use of Contrastive Predictive Coding features to improve automatic speaker verification systems by leveraging predictive coding and noise contrastive estimation techniques.

Contribution

It introduces CPC-based features into speaker verification, detailing methods, experiments, and analysis to enhance system performance.

Findings

CPC features improve speaker verification accuracy

Enhanced robustness to noise and variability

Demonstrated effectiveness over traditional features

Abstract

This thesis describes our ongoing work on Contrastive Predictive Coding (CPC) features for speaker verification. CPC is a recently proposed representation learning framework based on predictive coding and noise contrastive estimation. We focus on incorporating CPC features into the standard automatic speaker verification systems, and we present our methods, experiments, and analysis. This thesis also details necessary background knowledge in past and recent work on automatic speaker verification systems, conventional speech features, and the motivation and techniques behind CPC.