Adversarial Training for Multi-domain Speaker Recognition

by   Qing Wang, et al.

In real-life applications, the performance of speaker recognition systems always degrades when there is a mismatch between training and evaluation data. Many domain adaptation methods have been successfully used for eliminating the domain mismatches in speaker recognition. However, usually both training and evaluation data themselves can be composed of several subsets. These inner variances of each dataset can also be considered as different domains. Different distributed subsets in source or target domain dataset can also cause multi-domain mismatches, which are influential to speaker recognition performance. In this study, we propose to use adversarial training for multi-domain speaker recognition to solve the domain mismatch and the dataset variance problems. By adopting the proposed method, we are able to obtain both multi-domain-invariant and speaker-discriminative speech representations for speaker recognition. Experimental results on DAC13 dataset indicate that the proposed method is not only effective to solve the multi-domain mismatch problem, but also outperforms the compared unsupervised domain adaptation methods.


page 1

page 2

page 3

page 4


The CORAL++ Algorithm for Unsupervised Domain Adaptation of Speaker Recogntion

State-of-the-art speaker recognition systems are trained with a large am...

Channel adversarial training for cross-channel text-independent speaker recognition

The conventional speaker recognition frameworks (e.g., the i-vector and ...

Autoencoder based Domain Adaptation for Speaker Recognition under Insufficient Channel Information

In real-life conditions, mismatch between development and test domain de...

Unsupervised Domain Adaptation for Dysarthric Speech Detection via Domain Adversarial Training and Mutual Information Minimization

Dysarthric speech detection (DSD) systems aim to detect characteristics ...

DEAAN: Disentangled Embedding and Adversarial Adaptation Network for Robust Speaker Representation Learning

Despite speaker verification has achieved significant performance improv...

Speaker verification using end-to-end adversarial language adaptation

In this paper we investigate the use of adversarial domain adaptation fo...

The HCCL System for the NIST SRE21

This paper describes the systems developed by the HCCL team for the NIST...