The HCCL System for the NIST SRE21

TL;DR

The HCCL system for NIST SRE21 combines advanced speaker embedding techniques with novel domain adaptation methods to address cross-channel and cross-linguistic challenges, achieving significant improvements in speaker recognition performance.

Contribution

The paper introduces a comprehensive system integrating circle loss, data adaptation, and domain adaptation techniques specifically designed for the challenging conditions of NIST SRE21.

Findings

Data adaptation methods improved performance by 15%.

Applying speech enhancement reduces cross-domain mismatch.

Score calibration remains a challenge due to overfitting issues.

Abstract

This paper describes the systems developed by the HCCL team for the NIST 2021 speaker recognition evaluation (NIST SRE21).We first explore various state-of-the-art speaker embedding extractors combined with a novel circle loss to obtain discriminative deep speaker embeddings. Considering that cross-channel and cross-linguistic speaker recognition are the key challenges of SRE21, we introduce several techniques to reduce the cross-domain mismatch. Specifically, Codec and speech enhancement are directly applied to the raw speech to eliminate the codecs and the environment noise mismatch. We denote the methods that work directly on speech to eliminate the relatively explicit mismatches collectively as data adaptation methods. Experiments show that data adaption methods achieve 15\% improvements over our baseline. Furthermore, some popular back-ends domain adaptation algorithms are deployed on speaker embeddings to alleviate speaker performance degradation caused by the implicit mismatch. Score calibration is a major failure for us in SRE21. The reason is that score calibration with too many parameters easily lead to overfitting problems.