How to Improve Your Speaker Embeddings Extractor in Generic Toolkits

Recently, speaker embeddings extracted with deep neural networks became the state-of-the-art method for speaker verification. In this paper we aim to facilitate its implementation on a more generic toolkit than Kaldi, which we anticipate to enable further improvements on the method. We examine sever...

Full description

Saved in:
Bibliographic Details
Main Authors: Zeinali, Hossein, Burget, Lukas, Rohdin, Johan, Stafylakis, Themos, Cernocky, Jan
Format: Conference Proceeding
Language:English
Subjects:
Deep neural network
Feature extraction
Kaldi
Network topology
Neural networks
speaker embedding
Task analysis
Tensorflow
Topology
Training
Training data
x-vector
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Recently, speaker embeddings extracted with deep neural networks became the state-of-the-art method for speaker verification. In this paper we aim to facilitate its implementation on a more generic toolkit than Kaldi, which we anticipate to enable further improvements on the method. We examine several tricks in training, such as the effects of normalizing input features and pooled statistics, different methods for preventing overfitting as well as alternative non-linearities that can be used instead of Rectifier Linear Units. In addition, we investigate the difference in performance between TDNN and CNN, and between two types of attention mechanism. Experimental results on Speaker in the Wild, SRE 2016 and SRE 2018 datasets demonstrate the effectiveness of the proposed implementation.
ISSN:2379-190X
DOI:10.1109/ICASSP.2019.8683445