Speaker Weighted Training of HMM Using Multiple Reference Speakers

Hiroaki HATTORI  Satoshi NAKAMURA  Kiyohiro SHIKANO  Shigeki SAGAYAMA  

IEICE TRANSACTIONS on Information and Systems   Vol.E76-D   No.2   pp.219-226
Publication Date: 1993/02/25
Online ISSN: 
Print ISSN: 0916-8532
Type of Manuscript: PAPER
Category: Speech Processing
speaker adaptation,  HMM,  multiple reference speaker,  speaker weight,  

Full Text: PDF(675.3KB)>>
Buy this Article

This paper proposes a new speaker adaptation method using a speaker weighting technique for multiple reference speaker training of a hidden Markov model (HMM). The proposed method considers the similarities between an input speaker and multiple reference speakers, and use the similarities to control the influence of the reference speakers upon HMM. The evaluation experiments were carried out through the/b, d, g, m, n, N/phoneme recognition task using 8 speakers. Average recognition rates were 68.0%, 66.4%, and 65.6% respectively for three test sets which have different speech styles. These were 4.8%, 8.8%, and 10.5% higher than the rates of the spectrum mapping method, and also 1.6%, 6.7%, and 8.2% higher than the rates of the multiple reference speaker training, the supplemented HMM. The evaluation experiments clarified the effectiveness of the proposed method.