Presentation 2000/1/21
COMPARISON OF FEATURES OF MEL-CEPSTRUM AND SPECTRUM VECTOR FIELDS IN PHONEME RECOGNITION BASED ON THE BAYES DISCRIMINATION FUNCTION
Masayoshi Ihara, Takashi Akasaka, Ryuichi Oka,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper examines a method to label each frame of databases and voice waveforms, which are queries, in a voice search system. The phoneme recognition method is implemented by configuring a Bayes discrimination function for each phoneme category based on the feature information including two or more adjacent frames per phoname. This paper reports that almost the same recognition rates were obtained from comparative experiments on different, yet generally used, waveform analysis features:Mel-Ceptrum and spectrum vector fields.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Bayesian / Phoneme Recognition / Voice Recognition / Voice Retrieval
Paper # SP99-139
Date of Issue

Conference Information
Committee SP
Conference Date 2000/1/21(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 Speech (SP)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) COMPARISON OF FEATURES OF MEL-CEPSTRUM AND SPECTRUM VECTOR FIELDS IN PHONEME RECOGNITION BASED ON THE BAYES DISCRIMINATION FUNCTION
Sub Title (in English)
Keyword(1) Bayesian
Keyword(2) Phoneme Recognition
Keyword(3) Voice Recognition
Keyword(4) Voice Retrieval
1st Author's Name Masayoshi Ihara
1st Author's Affiliation Real World Computing Partnership, Tukuba Research Center()
2nd Author's Name Takashi Akasaka
2nd Author's Affiliation Real World Computing Partnership, Tukuba Research Center
3rd Author's Name Ryuichi Oka
3rd Author's Affiliation Real World Computing Partnership, Tsukuba Research Center
Date 2000/1/21
Paper # SP99-139
Volume (vol) vol.99
Number (no) 577
Page pp.pp.-
#Pages 8
Date of Issue