Serveur d'exploration sur la recherche en informatique en Lorraine

Attention, ce site est en cours de développement !
Attention, site généré par des moyens informatiques à partir de corpus bruts.
Les informations ne sont donc pas validées.

Acoustic and perceptual study of French stop bursts: Implications for stop recognition

Identifieur interne : 000D12 ( PascalFrancis/Checkpoint ); précédent : 000D11; suivant : 000D13

Acoustic and perceptual study of French stop bursts: Implications for stop recognition

Auteurs : Linda Djezzar [France]

Source :

RBID : Pascal:95-0557006

Descripteurs français

English descriptors

Abstract

A perceptual and acoustic investigation was conducted to better understand the discrimination power of the burst regarding place-of-articulation of French stops. These perceptual experiments showed that the burst provided very reliable spectral information about stop place and that prior knowledge of the vowel identity led to a slight but significant improvement in stop identification [A. Bonneau et al., to appear in J. Acoust. Soc. Am. (1995)]. This acoustic study confirmed that burst cues cannot be extracted and exploited without taking into account the vocalic context. Based on these conclusions, a hybrid recognizer was implemented. First, it recognizes the vocalic context, and then it takes this into account to extract and exploit the burst cues for the stop recognition task. The vocalic context was correctly recognized in 90% of the cases: Back vowels were perfectly recognized, central vowels with 80% accuracy, rounded front vowels with 90%, and unrounded front vowels with 89%. The global recognition rate of stop place was 88%. Statistical tests indicated that the best recognized stop was /k/ (93%), then /t/ (87%), and /p/ (84%). These results agree with listeners' performance observed in the perceptual experiments (/k/: 94% > /t/: 91% > /p/: 62%).


Affiliations:


Links toward previous steps (curation, corpus...)


Links to Exploration step

Pascal:95-0557006

Le document en format XML

<record>
<TEI>
<teiHeader>
<fileDesc>
<titleStmt>
<title xml:lang="en" level="a">Acoustic and perceptual study of French stop bursts: Implications for stop recognition</title>
<author>
<name sortKey="Djezzar, Linda" sort="Djezzar, Linda" uniqKey="Djezzar L" first="Linda" last="Djezzar">Linda Djezzar</name>
<affiliation wicri:level="3">
<inist:fA14 i1="01">
<s1>CRIN-CNRS & INRIA Lorraine, BP 239, 54506 Vandœuvre-les-Nancy Cedex, France</s1>
<sZ>1 aut.</sZ>
</inist:fA14>
<country xml:lang="fr">France</country>
<wicri:regionArea>CRIN-CNRS & INRIA Lorraine, BP 239, 54506 Vandœuvre-les-Nancy Cedex</wicri:regionArea>
<placeName>
<region type="region" nuts="2">Grand Est</region>
<region type="old region" nuts="2">Lorraine (région)</region>
<settlement type="city">Vandœuvre-lès-Nancy</settlement>
</placeName>
</affiliation>
</author>
</titleStmt>
<publicationStmt>
<idno type="wicri:source">INIST</idno>
<idno type="inist">95-0557006</idno>
<date when="1995-11">1995-11</date>
<idno type="stanalyst">PASCAL 95-0557006 AIP</idno>
<idno type="RBID">Pascal:95-0557006</idno>
<idno type="wicri:Area/PascalFrancis/Corpus">000D53</idno>
<idno type="wicri:Area/PascalFrancis/Curation">000B41</idno>
<idno type="wicri:Area/PascalFrancis/Checkpoint">000D12</idno>
<idno type="wicri:explorRef" wicri:stream="PascalFrancis" wicri:step="Checkpoint">000D12</idno>
</publicationStmt>
<sourceDesc>
<biblStruct>
<analytic>
<title xml:lang="en" level="a">Acoustic and perceptual study of French stop bursts: Implications for stop recognition</title>
<author>
<name sortKey="Djezzar, Linda" sort="Djezzar, Linda" uniqKey="Djezzar L" first="Linda" last="Djezzar">Linda Djezzar</name>
<affiliation wicri:level="3">
<inist:fA14 i1="01">
<s1>CRIN-CNRS & INRIA Lorraine, BP 239, 54506 Vandœuvre-les-Nancy Cedex, France</s1>
<sZ>1 aut.</sZ>
</inist:fA14>
<country xml:lang="fr">France</country>
<wicri:regionArea>CRIN-CNRS & INRIA Lorraine, BP 239, 54506 Vandœuvre-les-Nancy Cedex</wicri:regionArea>
<placeName>
<region type="region" nuts="2">Grand Est</region>
<region type="old region" nuts="2">Lorraine (région)</region>
<settlement type="city">Vandœuvre-lès-Nancy</settlement>
</placeName>
</affiliation>
</author>
</analytic>
<series>
<title level="j" type="main">Journal of the Acoustical Society of America</title>
<title level="j" type="abbreviated">J. Acoust. Soc. Am.</title>
<idno type="ISSN">0001-4966</idno>
<imprint>
<date when="1995-11">1995-11</date>
</imprint>
</series>
</biblStruct>
</sourceDesc>
<seriesStmt>
<title level="j" type="main">Journal of the Acoustical Society of America</title>
<title level="j" type="abbreviated">J. Acoust. Soc. Am.</title>
<idno type="ISSN">0001-4966</idno>
</seriesStmt>
</fileDesc>
<profileDesc>
<textClass>
<keywords scheme="KwdEn" xml:lang="en">
<term>Acoustics</term>
<term>Experimental study</term>
<term>Statistics</term>
</keywords>
<keywords scheme="Pascal" xml:lang="fr">
<term>Etude expérimentale</term>
<term>4371H</term>
<term>4372N</term>
<term>Acoustique</term>
<term>Statistique</term>
</keywords>
<keywords scheme="Wicri" type="topic" xml:lang="fr">
<term>Acoustique</term>
<term>Statistique</term>
</keywords>
</textClass>
</profileDesc>
</teiHeader>
<front>
<div type="abstract" xml:lang="en">A perceptual and acoustic investigation was conducted to better understand the discrimination power of the burst regarding place-of-articulation of French stops. These perceptual experiments showed that the burst provided very reliable spectral information about stop place and that prior knowledge of the vowel identity led to a slight but significant improvement in stop identification [A. Bonneau et al., to appear in J. Acoust. Soc. Am. (1995)]. This acoustic study confirmed that burst cues cannot be extracted and exploited without taking into account the vocalic context. Based on these conclusions, a hybrid recognizer was implemented. First, it recognizes the vocalic context, and then it takes this into account to extract and exploit the burst cues for the stop recognition task. The vocalic context was correctly recognized in 90% of the cases: Back vowels were perfectly recognized, central vowels with 80% accuracy, rounded front vowels with 90%, and unrounded front vowels with 89%. The global recognition rate of stop place was 88%. Statistical tests indicated that the best recognized stop was /k/ (93%), then /t/ (87%), and /p/ (84%). These results agree with listeners' performance observed in the perceptual experiments (/k/: 94% > /t/: 91% > /p/: 62%).</div>
</front>
</TEI>
<inist>
<standard h6="B">
<pA>
<fA01 i1="01" i2="1">
<s0>0001-4966</s0>
</fA01>
<fA02 i1="01">
<s0>JASMAN</s0>
</fA02>
<fA03 i2="1">
<s0>J. Acoust. Soc. Am.</s0>
</fA03>
<fA05>
<s2>98</s2>
</fA05>
<fA06>
<s2>5</s2>
</fA06>
<fA08 i1="01" i2="1" l="ENG">
<s1>Acoustic and perceptual study of French stop bursts: Implications for stop recognition</s1>
</fA08>
<fA11 i1="01" i2="1">
<s1>DJEZZAR (Linda)</s1>
</fA11>
<fA14 i1="01">
<s1>CRIN-CNRS & INRIA Lorraine, BP 239, 54506 Vandœuvre-les-Nancy Cedex, France</s1>
<sZ>1 aut.</sZ>
</fA14>
<fA20>
<s1>2891-2891</s1>
</fA20>
<fA21>
<s1>1995-11</s1>
</fA21>
<fA23 i1="01">
<s0>ENG</s0>
</fA23>
<fA43 i1="01">
<s1>INIST</s1>
<s2>129</s2>
</fA43>
<fA44>
<s0>8100</s0>
<s1>© AIP</s1>
</fA44>
<fA47 i1="01" i2="1">
<s0>95-0557006</s0>
</fA47>
<fA60>
<s1>P</s1>
<s2>C</s2>
<s3>E</s3>
</fA60>
<fA61>
<s0>A</s0>
</fA61>
<fA64 i1="01" i2="1">
<s0>Journal of the Acoustical Society of America</s0>
</fA64>
<fA66 i1="01">
<s0>USA</s0>
</fA66>
<fC01 i1="01" l="ENG">
<s0>A perceptual and acoustic investigation was conducted to better understand the discrimination power of the burst regarding place-of-articulation of French stops. These perceptual experiments showed that the burst provided very reliable spectral information about stop place and that prior knowledge of the vowel identity led to a slight but significant improvement in stop identification [A. Bonneau et al., to appear in J. Acoust. Soc. Am. (1995)]. This acoustic study confirmed that burst cues cannot be extracted and exploited without taking into account the vocalic context. Based on these conclusions, a hybrid recognizer was implemented. First, it recognizes the vocalic context, and then it takes this into account to extract and exploit the burst cues for the stop recognition task. The vocalic context was correctly recognized in 90% of the cases: Back vowels were perfectly recognized, central vowels with 80% accuracy, rounded front vowels with 90%, and unrounded front vowels with 89%. The global recognition rate of stop place was 88%. Statistical tests indicated that the best recognized stop was /k/ (93%), then /t/ (87%), and /p/ (84%). These results agree with listeners' performance observed in the perceptual experiments (/k/: 94% > /t/: 91% > /p/: 62%).</s0>
</fC01>
<fC02 i1="01" i2="X">
<s0>002A26I03</s0>
</fC02>
<fC02 i1="02" i2="1">
<s0>001D04A05B</s0>
</fC02>
<fC03 i1="01" i2="3" l="FRE">
<s0>Etude expérimentale</s0>
</fC03>
<fC03 i1="01" i2="3" l="ENG">
<s0>Experimental study</s0>
</fC03>
<fC03 i1="02" i2="3" l="FRE">
<s0>4371H</s0>
<s2>PAC</s2>
<s4>INC</s4>
</fC03>
<fC03 i1="03" i2="3" l="FRE">
<s0>4372N</s0>
<s2>PAC</s2>
<s4>INC</s4>
</fC03>
<fC03 i1="04" i2="3" l="FRE">
<s0>Acoustique</s0>
</fC03>
<fC03 i1="04" i2="3" l="ENG">
<s0>Acoustics</s0>
</fC03>
<fC03 i1="05" i2="3" l="FRE">
<s0>Statistique</s0>
</fC03>
<fC03 i1="05" i2="3" l="ENG">
<s0>Statistics</s0>
</fC03>
<fC07 i1="01" i2="3" l="FRE">
<s0>Sursaut</s0>
</fC07>
<fC07 i1="01" i2="3" l="ENG">
<s0>Bursts</s0>
</fC07>
<fC07 i1="02" i2="3" l="FRE">
<s0>Analyse domaine fréquence</s0>
</fC07>
<fC07 i1="02" i2="3" l="ENG">
<s0>Frequency domain analysis</s0>
</fC07>
<fC07 i1="03" i2="3" l="FRE">
<s0>Reconnaissance parole</s0>
</fC07>
<fC07 i1="03" i2="3" l="ENG">
<s0>Speech recognition</s0>
</fC07>
<fC07 i1="04" i2="3" l="FRE">
<s0>Voyelle</s0>
</fC07>
<fC07 i1="04" i2="3" l="ENG">
<s0>Vowels</s0>
</fC07>
<fN21>
<s1>317</s1>
</fN21>
<fN47 i1="01" i2="1">
<s0>9521M0799</s0>
</fN47>
</pA>
<pR>
<fA30 i1="01" i2="1" l="ENG">
<s1>The 130th meeting of the Acoustical Society of America</s1>
<s3>St. Louis, Missouri (USA)</s3>
<s4>1995-11-27/1995-12-01</s4>
</fA30>
</pR>
</standard>
</inist>
<affiliations>
<list>
<country>
<li>France</li>
</country>
<region>
<li>Grand Est</li>
<li>Lorraine (région)</li>
</region>
<settlement>
<li>Vandœuvre-lès-Nancy</li>
</settlement>
</list>
<tree>
<country name="France">
<region name="Grand Est">
<name sortKey="Djezzar, Linda" sort="Djezzar, Linda" uniqKey="Djezzar L" first="Linda" last="Djezzar">Linda Djezzar</name>
</region>
</country>
</tree>
</affiliations>
</record>

Pour manipuler ce document sous Unix (Dilib)

EXPLOR_STEP=$WICRI_ROOT/Wicri/Lorraine/explor/InforLorV4/Data/PascalFrancis/Checkpoint
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 000D12 | SxmlIndent | more

Ou

HfdSelect -h $EXPLOR_AREA/Data/PascalFrancis/Checkpoint/biblio.hfd -nk 000D12 | SxmlIndent | more

Pour mettre un lien sur cette page dans le réseau Wicri

{{Explor lien
   |wiki=    Wicri/Lorraine
   |area=    InforLorV4
   |flux=    PascalFrancis
   |étape=   Checkpoint
   |type=    RBID
   |clé=     Pascal:95-0557006
   |texte=   Acoustic and perceptual study of French stop bursts: Implications for stop recognition
}}

Wicri

This area was generated with Dilib version V0.6.33.
Data generation: Mon Jun 10 21:56:28 2019. Site generation: Fri Feb 25 15:29:27 2022