Seminaire LIA: Natalia Tomashenko

Title: Speaker adaptation of DNN acoustic models using Gaussian mixture model framework in ASR systems
Location: LIA, Salle de réunion
Speaker: Natalia Tomashenko
Date: May 24th, 2018 at 15h-16h
Summary: Adaptation is an efficient way to reduce mismatches between models and data from a particular speaker or channel in automatic speech recognition (ASR) systems. In this work, we present a novel speaker adaptation method for deep neural network (DNN) acoustic models. The idea of the proposed approach is based on using so-called GMM-derived features as input to a DNN. This technique of processing features for DNNs makes it possible to use GMM-HMM adaptation algorithms in the neural network framework. Adaptation to a new speaker can be performed by adapting an auxiliary GMM-HMM model used in calculation of GMM-derived features. The proposed approach is explored for various state-of-the art ASR systems and is shown to be effective in comparison with other speaker adaptation techniques and complementary to them.
Jeudi, 24 Mai, 2018 - 15:00 to 16:00

Laboratoire Informatique d'Avignon

Université d'Avignon et des Pays de Vaucluse
339 chemin des Meinajaries, Agroparc BP 91228, 84911 Avignon cedex 9
+33 (0)4 90 84 35 00