Clova Baseline System for the VoxCeleb Speaker Recognition Challenge 2020

TL;DR

This paper presents a ResNet-based speaker recognition system for VoxCeleb Challenge 2020, demonstrating significant improvements over existing methods without ensemble or post-processing, and providing open-source code and models.

Contribution

It introduces optimized ResNet variants trained with various loss functions, achieving state-of-the-art results as a baseline for the challenge.

Findings

Significant performance improvements over previous works

Effective ResNet variants trained with different loss functions

Open-source code and pre-trained models provided

Abstract

This report describes our submission to the VoxCeleb Speaker Recognition Challenge (VoxSRC) at Interspeech 2020. We perform a careful analysis of speaker recognition models based on the popular ResNet architecture, and train a number of variants using a range of loss functions. Our results show significant improvements over most existing works without the use of model ensemble or post-processing. We release the training code and pre-trained models as unofficial baselines for this year's challenge.