FORMATION OF PHONOLOGICAL CONCEPT STRUCTURES FROM SPOKEN WORD SAMPLES

Hiroaki Kojima; Kazuyo Tanaka; Satoru Hayamizu

FORMATION OF PHONOLOGICAL CONCEPT STRUCTURES FROM SPOKEN WORD SAMPLES

Hiroaki Kojima^*, Kazuyo Tanaka, Satoru Hayamizu

^*Corresponding author for this work

Research output: Contribution to conference › Paper › peer-review

5 Citations (Scopus)

Abstract

This paper describes a framework for phonological concept formation, which is the task of acquiring an efficient representation of phonemes from spoken word samples without using any transcriptions except for the identification of the words. The phoneme models are represented as networks of segments, each of which forms a compact, distribution of spectral features. We call this representation a phonological concept The learning process is performed by searching in a hypotheses space for which each hypothesis is produced by modifying a set of phoneme models. This system potentially enables us to improve speech recognition performance.

Original language	English
Pages	269-272
Number of pages	4
Publication status	Published - 1992
Externally published	Yes
Event	2nd International Conference on Spoken Language Processing, ICSLP 1992 - Banff, Canada Duration: 1992 Oct 13 → 1992 Oct 16

Conference

Conference	2nd International Conference on Spoken Language Processing, ICSLP 1992
Country/Territory	Canada
City	Banff
Period	92/10/13 → 92/10/16

ASJC Scopus subject areas

Language and Linguistics
Linguistics and Language

Cite this

@conference{3891a59c29354ab59442e75fbb96e045,

title = "FORMATION OF PHONOLOGICAL CONCEPT STRUCTURES FROM SPOKEN WORD SAMPLES",

abstract = "This paper describes a framework for phonological concept formation, which is the task of acquiring an efficient representation of phonemes from spoken word samples without using any transcriptions except for the identification of the words. The phoneme models are represented as networks of segments, each of which forms a compact, distribution of spectral features. We call this representation a phonological concept The learning process is performed by searching in a hypotheses space for which each hypothesis is produced by modifying a set of phoneme models. This system potentially enables us to improve speech recognition performance.",

author = "Hiroaki Kojima and Kazuyo Tanaka and Satoru Hayamizu",

note = "Publisher Copyright: {\textcopyright} 1992 2nd International Conference on Spoken Language Processing, ICSLP 1992. All rights reserved.; 2nd International Conference on Spoken Language Processing, ICSLP 1992 ; Conference date: 13-10-1992 Through 16-10-1992",

year = "1992",

language = "English",

pages = "269--272",

}

TY - CONF

T1 - FORMATION OF PHONOLOGICAL CONCEPT STRUCTURES FROM SPOKEN WORD SAMPLES

AU - Kojima, Hiroaki

AU - Tanaka, Kazuyo

AU - Hayamizu, Satoru

PY - 1992

Y1 - 1992

N2 - This paper describes a framework for phonological concept formation, which is the task of acquiring an efficient representation of phonemes from spoken word samples without using any transcriptions except for the identification of the words. The phoneme models are represented as networks of segments, each of which forms a compact, distribution of spectral features. We call this representation a phonological concept The learning process is performed by searching in a hypotheses space for which each hypothesis is produced by modifying a set of phoneme models. This system potentially enables us to improve speech recognition performance.

AB - This paper describes a framework for phonological concept formation, which is the task of acquiring an efficient representation of phonemes from spoken word samples without using any transcriptions except for the identification of the words. The phoneme models are represented as networks of segments, each of which forms a compact, distribution of spectral features. We call this representation a phonological concept The learning process is performed by searching in a hypotheses space for which each hypothesis is produced by modifying a set of phoneme models. This system potentially enables us to improve speech recognition performance.

UR - http://www.scopus.com/inward/record.url?scp=85009084813&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85009084813&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:85009084813

SP - 269

EP - 272

T2 - 2nd International Conference on Spoken Language Processing, ICSLP 1992

Y2 - 13 October 1992 through 16 October 1992

ER -

FORMATION OF PHONOLOGICAL CONCEPT STRUCTURES FROM SPOKEN WORD SAMPLES

Abstract

Conference

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this