Presentation 1996/12/12
Automatic Generation of Pronunciation Dictionary Based on Pronunciation Networks
Toshiaki Fukada, Yoshinori Sagisaka,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In spontaneous speech, word pronunciations vary more than in read speech, while only standard pronunciations in citation form are used in most Japanese recognition systems disregarding real pronunciation variations. To cope with pronunciation variations in spontaneous speech, it is important to build a pronunciation dictionary from real speech data. In this paper, we propose an automatic generation method of a pronunciation dictionary based on pronunciation networks which give the most plausible pronunciations (henceforth correct pronunciations). First, phoneme recognition results are taken as the correct pronunciations. Then pronunciation networks are trained using multi-layer perceptron neural networks to predict these correct pronunciations from the canonical symbol sequences. Finally, pronunciation dictionaries are automatically generated from the pronunciation networks. Experimental results on spontaneous speech show that the automatically-derived pronunciation dictionaries give consistently higher recognition rates and require less computational time for recognition than the conventional dictionary.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) pronunciation dictionary / neural networks / spontaneous speech / HMM / speech recognition
Paper # NLC96-40,SP96-71
Date of Issue

Conference Information
Committee NLC
Conference Date 1996/12/12(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Natural Language Understanding and Models of Communication (NLC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Automatic Generation of Pronunciation Dictionary Based on Pronunciation Networks
Sub Title (in English)
Keyword(1) pronunciation dictionary
Keyword(2) neural networks
Keyword(3) spontaneous speech
Keyword(4) HMM
Keyword(5) speech recognition
1st Author's Name Toshiaki Fukada
1st Author's Affiliation ATR Interpreting Telecommunications Research Laboratories()
2nd Author's Name Yoshinori Sagisaka
2nd Author's Affiliation ATR Interpreting Telecommunications Research Laboratories
Date 1996/12/12
Paper # NLC96-40,SP96-71
Volume (vol) vol.96
Number (no) 419
Page pp.pp.-
#Pages 8
Date of Issue