Presentation 1996/2/23
Automatic Acquisition of Speech Units for Ultra Very Low Bit Coding
Minoru SAITO, Mikio MASUKATA, Seiichi NAKAGAWA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) For ultra very low bit speech coding as less than 100bps, we describe the method to acquire speech units automatically which have almost same time length as a syllable. Optimal speech units are obtained with the procedure that, the number of units is given beforehand and pattern of units is updated successively based on a minimum distortion criterion. At this time, a restriction is embeded that utterances spoken as the same word are expressed the same sequences of speech units, so we present algorithms for this purpose. When we acquired speech units automatically on condition both the number and length of units equivalent to syllables and used these for 216 words recognition experiment, the recognition rate was obtained higher than that of syllable-based method. And the word recognition rate about, 98% or so was achieved by increasing the number of units, while the bit rate was still less than 60bps.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) ultra very low bit coding / automatic acquisition of speech units / speech recognition
Paper # SP95-132
Date of Issue

Conference Information
Committee SP
Conference Date 1996/2/23(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) Automatic Acquisition of Speech Units for Ultra Very Low Bit Coding
Sub Title (in English)
Keyword(1) ultra very low bit coding
Keyword(2) automatic acquisition of speech units
Keyword(3) speech recognition
1st Author's Name Minoru SAITO
1st Author's Affiliation Department of Information and Computer Sciences Toyohashi University of Technology()
2nd Author's Name Mikio MASUKATA
2nd Author's Affiliation Department of Information and Computer Sciences Toyohashi University of Technology
3rd Author's Name Seiichi NAKAGAWA
3rd Author's Affiliation Department of Information and Computer Sciences Toyohashi University of Technology
Date 1996/2/23
Paper # SP95-132
Volume (vol) vol.95
Number (no) 534
Page pp.pp.-
#Pages 8
Date of Issue