Unsupervised Speaker Adaptation Using Speaker-Class Models for Lecture Speech Recognition

Abstract

In this paper, we propose a new speaker-class modeling and its adaptation method for the LVCSR system and evaluate the method on the Corpus of Spontaneous Japanese (CSJ). In this method, closer speakers are selected from training speakers and the acoustic models are trained by using their utterances for each evaluation speaker. One of the major issues of the speaker-class model is determining the selection range of speakers. In order to solve the problem, several models which have a variety of speaker range are prepared for each evaluation speaker in advance, and the most proper model is selected on a likelihood basis in the recognition step. In addition, we improved the recognition performance using unsupervised speaker adaptation with the speaker-class models. In the recognition experiments, a significant improvement could be obtained by using the proposed speaker adaptation based on speaker-class models compared with the conventional adaptation method.

Related Papers

• Integrating Speaker Identification and Learning with Adaptive Speech Recognition(2004)
• → Speaker Segmentation and Adaptation for Speech Recognition on Multiple-Speaker Audio Conference Data(2007)6 cited
• → A speaker identification system with verification method based on speaker relative threshold and HMM(2003)4 cited
• → Speaker Recognition and Diarization(2010)3 cited
• → Factors responsible and phases of speaker recognition system(2022)