SPEAKER ADAPTATION FITTING TRAINING DATA SIZE AND CONTENTS

Masahiro Tonomura, Tetsuo Kosaka, Shoichi Matsunaga, Akito Monden

Research output: Contribution to conferencePaperpeer-review

2 Citations (Scopus)

Abstract

This paper proposes a speaker adaptation algorithm that covers a wide range of adaptation data. The parameter smoothing technique improves adaptation performance for a small amount of adaptation data; however, this smoothing usually reduces adaptation efficiency for a large amount of adaptation data. Our method dynamically controls smoothing strength by using information on the amount of adaptation data to achieve good adaptation performance over a wide range of adaptation data. The proposed method is combined with the maximum a posteriori (MAP) estimation technique, and its effectiveness is shown on a Japanese 26 phoneme recognition test.

Original languageEnglish
Pages1147-1150
Number of pages4
Publication statusPublished - 1995
Externally publishedYes
Event4th European Conference on Speech Communication and Technology, EUROSPEECH 1995 - Madrid, Spain
Duration: Sep 18 1995Sep 21 1995

Conference

Conference4th European Conference on Speech Communication and Technology, EUROSPEECH 1995
Country/TerritorySpain
CityMadrid
Period9/18/959/21/95

ASJC Scopus subject areas

  • Computer Science Applications
  • Software
  • Linguistics and Language
  • Communication

Fingerprint

Dive into the research topics of 'SPEAKER ADAPTATION FITTING TRAINING DATA SIZE AND CONTENTS'. Together they form a unique fingerprint.

Cite this