AUTOMATIC LABELING OF KNOWN SPEECH SAMPLES USING A RULE-BASED NETWORK REPRESENTATION AND SEGMENTATION TECHNIQUE.

Kazuyo Tanaka; Satoru Hayamizu; Kozo Ohta

AUTOMATIC LABELING OF KNOWN SPEECH SAMPLES USING A RULE-BASED NETWORK REPRESENTATION AND SEGMENTATION TECHNIQUE.

Kazuyo Tanaka^*, Satoru Hayamizu, Kozo Ohta

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

Abstract

An automatic labeling technique for known speech samples is proposed to construct a fine speech data base. A word (or sentence) is represented by a phonetic network which covers the acoustic variation contained in the utterances of the word (or sentence). An input speech sample is segmented using its parameter pattern dynamics and labeled to the optimal phonetic label (called APSEG) sequence by matching th segment sequence to the generated phonetic network using constrained dynamic programming. The feasibility of the method is confirmed when it is applied ot a word set containing 53 city names.

Original language	English
Pages (from-to)	30-37
Number of pages	8
Journal	Denshi Gijutsu Sogo Kenkyusho Iho/Bulletin of the Electrotechnical Laboratory
Volume	52
Issue number	3
Publication status	Published - 1988
Externally published	Yes

ASJC Scopus subject areas

Condensed Matter Physics
Electrical and Electronic Engineering

Cite this

@article{eb66ce978ab54ee285ea2fc8cbef0320,

title = "AUTOMATIC LABELING OF KNOWN SPEECH SAMPLES USING A RULE-BASED NETWORK REPRESENTATION AND SEGMENTATION TECHNIQUE.",

abstract = "An automatic labeling technique for known speech samples is proposed to construct a fine speech data base. A word (or sentence) is represented by a phonetic network which covers the acoustic variation contained in the utterances of the word (or sentence). An input speech sample is segmented using its parameter pattern dynamics and labeled to the optimal phonetic label (called APSEG) sequence by matching th segment sequence to the generated phonetic network using constrained dynamic programming. The feasibility of the method is confirmed when it is applied ot a word set containing 53 city names.",

author = "Kazuyo Tanaka and Satoru Hayamizu and Kozo Ohta",

year = "1988",

language = "English",

volume = "52",

pages = "30--37",

journal = "Denshi Gijutsu Sogo Kenkyusho Iho/Bulletin of the Electrotechnical Laboratory",

issn = "0366-9092",

publisher = "Denshi Gijutsu Sogo Kenkyujo",

number = "3",

}

TY - JOUR

T1 - AUTOMATIC LABELING OF KNOWN SPEECH SAMPLES USING A RULE-BASED NETWORK REPRESENTATION AND SEGMENTATION TECHNIQUE.

AU - Tanaka, Kazuyo

AU - Hayamizu, Satoru

AU - Ohta, Kozo

PY - 1988

Y1 - 1988

N2 - An automatic labeling technique for known speech samples is proposed to construct a fine speech data base. A word (or sentence) is represented by a phonetic network which covers the acoustic variation contained in the utterances of the word (or sentence). An input speech sample is segmented using its parameter pattern dynamics and labeled to the optimal phonetic label (called APSEG) sequence by matching th segment sequence to the generated phonetic network using constrained dynamic programming. The feasibility of the method is confirmed when it is applied ot a word set containing 53 city names.

AB - An automatic labeling technique for known speech samples is proposed to construct a fine speech data base. A word (or sentence) is represented by a phonetic network which covers the acoustic variation contained in the utterances of the word (or sentence). An input speech sample is segmented using its parameter pattern dynamics and labeled to the optimal phonetic label (called APSEG) sequence by matching th segment sequence to the generated phonetic network using constrained dynamic programming. The feasibility of the method is confirmed when it is applied ot a word set containing 53 city names.

UR - http://www.scopus.com/inward/record.url?scp=0023859810&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0023859810&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0023859810

SN - 0366-9092

VL - 52

SP - 30

EP - 37

JO - Denshi Gijutsu Sogo Kenkyusho Iho/Bulletin of the Electrotechnical Laboratory

JF - Denshi Gijutsu Sogo Kenkyusho Iho/Bulletin of the Electrotechnical Laboratory

IS - 3

ER -

AUTOMATIC LABELING OF KNOWN SPEECH SAMPLES USING A RULE-BASED NETWORK REPRESENTATION AND SEGMENTATION TECHNIQUE.

Abstract

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this