Exploring Gaussian mixture model framework for speaker adaptation of deep neural network acoustic models

by   Natalia Tomashenko, et al.

In this paper we investigate the GMM-derived (GMMD) features for adaptation of deep neural network (DNN) acoustic models. The adaptation of the DNN trained on GMMD features is done through the maximum a posteriori (MAP) adaptation of the auxiliary GMM model used for GMMD feature extraction. We explore fusion of the adapted GMMD features with conventional features, such as bottleneck and MFCC features, in two different neural network architectures: DNN and time-delay neural network (TDNN). We analyze and compare different types of adaptation techniques such as i-vectors and feature-space adaptation techniques based on maximum likelihood linear regression (fMLLR) with the proposed adaptation approach, and explore their complementarity using various types of fusion such as feature level, posterior level, lattice level and others in order to discover the best possible way of combination. Experimental results on the TED-LIUM corpus show that the proposed adaptation technique can be effectively integrated into DNN and TDNN setups at different levels and provide additional gain in recognition performance: up to 6 rate reduction (WERR) over the strong feature-space adaptation techniques based on maximum likelihood linear regression (fMLLR) speaker adapted DNN baseline, and up to 18 DNN baseline model, trained on conventional features. For TDNN models the proposed approach achieves up to 26 baseline, and up 13 The analysis of the adapted GMMD features from various points of view demonstrates their effectiveness at different levels.


page 1

page 2

page 3

page 4


Maximum a Posteriori Adaptation of Network Parameters in Deep Models

We present a Bayesian approach to adapting parameters of a well-trained ...

Adversarial Speaker Adaptation

We propose a novel adversarial speaker adaptation (ASA) scheme, in which...

Integration of TensorFlow based Acoustic Model with Kaldi WFST Decoder

While the Kaldi framework provides state-of-the-art components for speec...

Empirical Evaluation of Speaker Adaptation on DNN based Acoustic Model

Speaker adaptation aims to estimate a speaker specific acoustic model fr...

Differentiable Pooling for Unsupervised Acoustic Model Adaptation

We present a deep neural network (DNN) acoustic model that includes para...

Statistical Models in Forensic Voice Comparison

This chapter describes a number of signal-processing and statistical-mod...

A Comparative Re-Assessment of Feature Extractors for Deep Speaker Embeddings

Modern automatic speaker verification relies largely on deep neural netw...