DeepAI
Log In Sign Up

Learning Metrics from Mean Teacher: A Supervised Learning Method for Improving the Generalization of Speaker Verification System

04/14/2021
by   Ju-ho Kim, et al.
0

Most speaker verification tasks are studied as an open-set evaluation scenario considering the real-world condition. Thus, the generalization power to unseen speakers is of paramount important to the performance of the speaker verification system. We propose to apply Mean Teacher, a temporal averaging model, to extract speaker embeddings with small intra-class variance and large inter-class variance. The mean teacher network refers to the temporal averaging of deep neural network parameters; it can produces more accurate and stable representations than using weights after the training finished. By learning the reliable intermediate representation of the mean teacher network, we expect that the proposed method can explore more discriminatory embedding spaces and improve the generalization performance of the speaker verification system. Experimental results on the VoxCeleb1 test set demonstrate that the proposed method relatively improves performance by 11.61%, compared to a baseline system.

READ FULL TEXT

page 1

page 2

page 3

page 4

05/07/2020

Crop Aggregating for short utterances speaker verification using raw waveforms

Most studies on speaker verification systems focus on long-duration utte...
05/07/2020

Segment Aggregation for short utterances speaker verification using raw waveforms

Most studies on speaker verification systems focus on long-duration utte...
03/20/2020

Improving Embedding Extraction for Speaker Verification with Ladder Network

Speaker verification is an established yet challenging task in speech pr...
01/28/2020

Masked cross self-attention encoding for deep speaker embedding

In general, speaker verification tasks require the extraction of speaker...
11/24/2021

An MAP Estimation for Between-Class Variance

Probabilistic linear discriminant analysis (PLDA) has been widely used i...
07/13/2022

Cross-Age Speaker Verification: Learning Age-Invariant Speaker Embeddings

Automatic speaker verification has achieved remarkable progress in recen...
06/17/2021

Multi-Level Transfer Learning from Near-Field to Far-Field Speaker Verification

In far-field speaker verification, the performance of speaker embeddings...