<?xml version="1.0" encoding="ISO-8859-1"?><article xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
<front>
<journal-meta>
<journal-id>1405-5546</journal-id>
<journal-title><![CDATA[Computación y Sistemas]]></journal-title>
<abbrev-journal-title><![CDATA[Comp. y Sist.]]></abbrev-journal-title>
<issn>1405-5546</issn>
<publisher>
<publisher-name><![CDATA[Instituto Politécnico Nacional, Centro de Investigación en Computación]]></publisher-name>
</publisher>
</journal-meta>
<article-meta>
<article-id>S1405-55462003000400006</article-id>
<title-group>
<article-title xml:lang="en"><![CDATA[Face Recognition Using Unlabeled Data]]></article-title>
<article-title xml:lang="es"><![CDATA[Reconocimiento de Rostros usando Datos No Etiquetados]]></article-title>
</title-group>
<contrib-group>
<contrib contrib-type="author">
<name>
<surname><![CDATA[Martínez]]></surname>
<given-names><![CDATA[Carmen]]></given-names>
</name>
<xref ref-type="aff" rid="A01"/>
</contrib>
<contrib contrib-type="author">
<name>
<surname><![CDATA[Fuentes]]></surname>
<given-names><![CDATA[Olac]]></given-names>
</name>
<xref ref-type="aff" rid="A01"/>
</contrib>
</contrib-group>
<aff id="A01">
<institution><![CDATA[,Instituto Nacional de Astrofísica, Óptica y Electrónica  ]]></institution>
<addr-line><![CDATA[ Puebla]]></addr-line>
<country>México</country>
</aff>
<pub-date pub-type="pub">
<day>00</day>
<month>12</month>
<year>2003</year>
</pub-date>
<pub-date pub-type="epub">
<day>00</day>
<month>12</month>
<year>2003</year>
</pub-date>
<volume>7</volume>
<numero>2</numero>
<fpage>123</fpage>
<lpage>129</lpage>
<copyright-statement/>
<copyright-year/>
<self-uri xlink:href="http://www.scielo.org.mx/scielo.php?script=sci_arttext&amp;pid=S1405-55462003000400006&amp;lng=en&amp;nrm=iso"></self-uri><self-uri xlink:href="http://www.scielo.org.mx/scielo.php?script=sci_abstract&amp;pid=S1405-55462003000400006&amp;lng=en&amp;nrm=iso"></self-uri><self-uri xlink:href="http://www.scielo.org.mx/scielo.php?script=sci_pdf&amp;pid=S1405-55462003000400006&amp;lng=en&amp;nrm=iso"></self-uri><abstract abstract-type="short" xml:lang="en"><p><![CDATA[Face recognition systems can normally attain good accuracy when they are provided with a large set of training examples. However, when a large training set is not available, their performance is commonly poor. In this work we describe a method for face recognition that achieves good results when only a very small training set is available (one image per person). The method is based on augmenting the original training set with previously unlabeled data (that is, face images for which the identity of the person is not known). Initially, we apply the well-known eigenfaces technique to reduce the dimensionality of the image space, then we perform an iterative process, classifying all the unlabeled data with an ensemble of classifiers built from the current training set, and appending to the training set the previously unlabeled examples that are believed to be correctly classified with a high confidence level, according to the ensemble. We experimented with ensembles based on the k-nearest neighbors, feed forward artificial neural networks and locally weighted linear regression learning algorithms. Our experimental results show that using unlabeled data improves the accuracy in all cases. The best accuracy, 92.07%, was obtained with locally weighted linear regression using 30 eigenfaces and appending 3 examples of every class in each iteration. In contrast, using only labeled data, an accuracy of only 34.81% was obtained.]]></p></abstract>
<abstract abstract-type="short" xml:lang="es"><p><![CDATA[Los sistemas de reconocimiento de rostros normalmente obtienen buenos resultados cuando tienen disponibles conjuntos de entrenamiento grandes. Sin embargo, cuando no hay un conjunto de entrenamiento grande disponible, su desempeño no es satisfactorio. En este trabajo presentamos un método para reconocimiento de rostros que obtiene buenos resultados cuando solo se tiene disponible un conjunto de entrenamiento pequeño (incluso una sola imagen por persona). El método se basa en expandir el conjunto de entrenamiento original usando datos no etiquetados previamente (esto es, imágenes de rostros con identidad desconocida). Inicialmente, aplicamos la técnica de eigenrostros para reducir la dimensionalidad del espacio de atributos, después realizamos un proceso iterativo, clasificando todos los datos no etiquetados con un ensamble de clasificadores construido a partir del conjunto de entrenamiento actual y agregando al conjunto de entrenamiento los ejemplos que han sido clasificados correctamente con un alto nivel de confianza, de acuerdo al ensamble. Realizamos experimentos usando ensambles basados en el algoritmo de k vecinos más cercanos, redes neuronales artificiales, y regresión lineal localmente ponderada. Los resultados experimentales demuestran que el uso de datos no etiquetados mejora la clasificación en todos los casos. Los mejores resultados, con un porcentaje de clasificación correcta de 92.07, fueron obtenidos con regresión lineal localmente ponderada usando 30 eigenrostros y agregando 3 ejemplos de cada clase en cada iteración. Como comparación, usando únicamente los datos etiquetados, solo se clasificaron correctamente el 34.81% de los ejemplos.]]></p></abstract>
</article-meta>
</front><body><![CDATA[ <p align="justify"><font face="verdana" size="4">Art&iacute;culo</font></p>     <p align="center"><font face="verdana" size="2">&nbsp;</font></p>     <p align="center"><font face="verdana" size="4"><b>Face Recognition Using Unlabeled Data</b></font></p>     <p align="center"><font face="verdana" size="2">&nbsp;</font></p>     <p align="center"><font face="verdana" size="3"><b><i>Reconocimiento de Rostros usando Datos No Etiquetados</i></b></font></p>     <p align="center"><font face="verdana" size="2">&nbsp;</font></p>     <p align="center"><font face="verdana" size="2"><b>Carmen Mart&iacute;nez and Olac Fuentes</b></font></p>     <p align="justify"><font face="verdana" size="2">&nbsp;</font></p>     <p align="justify"><font face="verdana" size="2"><i>Instituto Nacional de Astrof&iacute;sica, &Oacute;ptica y Electr&oacute;nica Luis Enrique Erro # 1 Santa Maria Tonanzintla, Puebla, 72840, M&eacute;xico. </i>E&#150;mails: <i><a href="mailto:carmen@ccc.inaoep.mx">carmen@ccc.inaoep.mx</a> ; <a href="mailto:fuentes@inaoep.mx"> fuentes@inaoep.mx</a></i></font></p>     <p align="justify"><font face="verdana" size="2">&nbsp;</font></p>     ]]></body>
<body><![CDATA[<p align="justify"><font face="verdana" size="2"><b>Abstract</b></font></p>     <p align="justify"><font face="verdana" size="2">Face recognition systems can normally attain good accuracy when they are provided with a large set of training examples. However, when a large training set is not available, their performance is commonly poor. In this work we describe a method for face recognition that achieves good results when only a very small training set is available (one image per person). The method is based on augmenting the original training set with previously unlabeled data (that is, face images for which the identity of the person is not known). Initially, we apply the well&#150;known eigenfaces technique to reduce the dimensionality of the image space, then we perform an iterative process, classifying all the unlabeled data with an ensemble of classifiers built from the current training set, and appending to the training set the previously unlabeled examples that are believed to be correctly classified with a high confidence level, according to the ensemble.</font></p>     <p align="justify"><font face="verdana" size="2">We experimented with ensembles based on the k&#150;nearest neighbors, feed forward artificial neural networks and locally weighted linear regression learning algorithms. Our experimental results show that using unlabeled data improves the accuracy in all cases. The best accuracy, 92.07%, was obtained with locally weighted linear regression using 30 eigenfaces and appending 3 examples of every class in each iteration. In contrast, using only labeled data, an accuracy of only 34.81% was obtained.</font></p>     <p align="justify"><font face="verdana" size="2">&nbsp;</font></p>     <p align="justify"><font face="verdana" size="2"><b>Resumen</b></font></p>     <p align="justify"><font face="verdana" size="2">Los sistemas de reconocimiento de rostros normalmente obtienen buenos resultados cuando tienen disponibles conjuntos de entrenamiento grandes. Sin embargo, cuando no hay un conjunto de entrenamiento grande disponible, su desempe&ntilde;o no es satisfactorio. En este trabajo presentamos un m&eacute;todo para reconocimiento de rostros que obtiene buenos resultados cuando solo se tiene disponible un conjunto de entrenamiento peque&ntilde;o (incluso una sola imagen por persona). El m&eacute;todo se basa en expandir el conjunto de entrenamiento original usando datos no etiquetados previamente (esto es, im&aacute;genes de rostros con identidad desconocida). Inicialmente, aplicamos la t&eacute;cnica de eigenrostros para reducir la dimensionalidad del espacio de atributos, despu&eacute;s realizamos un proceso iterativo, clasificando todos los datos no etiquetados con un ensamble de clasificadores construido a partir del conjunto de entrenamiento actual y agregando al conjunto de entrenamiento los ejemplos que han sido clasificados correctamente con un alto nivel de confianza, de acuerdo al ensamble.</font></p>     <p align="justify"><font face="verdana" size="2">Realizamos experimentos usando ensambles basados en el algoritmo de k vecinos m&aacute;s cercanos, redes neuronales artificiales, y regresi&oacute;n lineal localmente ponderada. Los resultados experimentales demuestran que el uso de datos no etiquetados mejora la clasificaci&oacute;n en todos los casos. Los mejores resultados, con un porcentaje de clasificaci&oacute;n correcta de 92.07, fueron obtenidos con regresi&oacute;n lineal localmente ponderada usando 30 eigenrostros y agregando 3 ejemplos de cada clase en cada iteraci&oacute;n. Como comparaci&oacute;n, usando &uacute;nicamente los datos etiquetados, solo se clasificaron correctamente el 34.81% de los ejemplos.</font></p>     <p align="justify"><font face="verdana" size="2">&nbsp;</font></p>     <p align="justify"><font face="verdana" size="2"><a href="/pdf/cys/v7n2/v7n2a6.pdf" target="_blank">DESCARGAR ART&Iacute;CULO EN FORMATO PDF</a></font></p>     <p align="justify"><font face="verdana" size="2">&nbsp;</font></p>     ]]></body>
<body><![CDATA[<p align="justify"><font face="verdana" size="2"><b>References</b></font></p>     <!-- ref --><p align="justify"><font face="verdana" size="2">1. <b>C. M. Bishop. </b>Neural Networks for Pattern Recognition. Oxford Universiy Press, Oxford, England, 1996.</font>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[&#160;<a href="javascript:void(0);" onclick="javascript: window.open('/scielo.php?script=sci_nlinks&ref=2046328&pid=S1405-5546200300040000600001&lng=','','width=640,height=500,resizable=yes,scrollbars=1,menubar=yes,');">Links</a>&#160;]<!-- end-ref --><!-- ref --><p align="justify"><font face="verdana" size="2">2. <b>A. Blum </b>and<b> T. Mitchell. </b>Learning to classify text from labeled and unlabeled documents. Conference on Computational Learning Theory, 1998.</font>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[&#160;<a href="javascript:void(0);" onclick="javascript: window.open('/scielo.php?script=sci_nlinks&ref=2046329&pid=S1405-5546200300040000600002&lng=','','width=640,height=500,resizable=yes,scrollbars=1,menubar=yes,');">Links</a>&#160;]<!-- end-ref --><!-- ref --><p align="justify"><font face="verdana" size="2">3. <b>T. G. Dietterich. </b>Machine learning research: Four current directions. The AI Magazine, 1997. </font>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[&#160;<a href="javascript:void(0);" onclick="javascript: window.open('/scielo.php?script=sci_nlinks&ref=2046330&pid=S1405-5546200300040000600003&lng=','','width=640,height=500,resizable=yes,scrollbars=1,menubar=yes,');">Links</a>&#160;]<!-- end-ref --><!-- ref --><p align="justify"><font face="verdana" size="2">4. <b>L. Fausett. </b>Fundamentals of Neural Networks: Architectures, Algorithms and Aplications. Prentice&#150; Hall, 1994. </font>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[&#160;<a href="javascript:void(0);" onclick="javascript: window.open('/scielo.php?script=sci_nlinks&ref=2046331&pid=S1405-5546200300040000600004&lng=','','width=640,height=500,resizable=yes,scrollbars=1,menubar=yes,');">Links</a>&#160;]<!-- end-ref --><!-- ref --><p align="justify"><font face="verdana" size="2">5. <b>Y. Freund, H. Seung, E. Shamir, </b>and<b> N. Tishby. </b>Selective sampling using the query by committee algorithm. Machine Learning, 1997. </font>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[&#160;<a href="javascript:void(0);" onclick="javascript: window.open('/scielo.php?script=sci_nlinks&ref=2046332&pid=S1405-5546200300040000600005&lng=','','width=640,height=500,resizable=yes,scrollbars=1,menubar=yes,');">Links</a>&#160;]<!-- end-ref --><!-- ref --><p align="justify"><font face="verdana" size="2">6. <b>D. Graham </b>and<b> N. Allinson. </b>Character&#150;rising Virtual Eigensignatures for Face Recognition. In Face Recognition: From Theory to Applications. Springer&#150;Verlag, 1998. </font>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[&#160;<a href="javascript:void(0);" onclick="javascript: window.open('/scielo.php?script=sci_nlinks&ref=2046333&pid=S1405-5546200300040000600006&lng=','','width=640,height=500,resizable=yes,scrollbars=1,menubar=yes,');">Links</a>&#160;]<!-- end-ref --><!-- ref --><p align="justify"><font face="verdana" size="2">7. <b>A. McCallum </b>and<b> K. Nigam. </b>Employing EM in pool&#150;based active learning for text classification. In Proceeding of the 15th International Conference on Machine Learning, 1998. </font>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[&#160;<a href="javascript:void(0);" onclick="javascript: window.open('/scielo.php?script=sci_nlinks&ref=2046334&pid=S1405-5546200300040000600007&lng=','','width=640,height=500,resizable=yes,scrollbars=1,menubar=yes,');">Links</a>&#160;]<!-- end-ref --><!-- ref --><p align="justify"><font face="verdana" size="2">8. <b>T. M. Mitchell. </b>Machine Learning. McGraw&#150;Hill, 1997. </font>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[&#160;<a href="javascript:void(0);" onclick="javascript: window.open('/scielo.php?script=sci_nlinks&ref=2046335&pid=S1405-5546200300040000600008&lng=','','width=640,height=500,resizable=yes,scrollbars=1,menubar=yes,');">Links</a>&#160;]<!-- end-ref --><!-- ref --><p align="justify"><font face="verdana" size="2">9. <b>K. Nigam, A. McCallum, S. Thrun, </b>and<b> T. Mitchell. </b>Learning to classify text from labeled and unlabeled documents. Machine Learning, 1999. </font>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[&#160;<a href="javascript:void(0);" onclick="javascript: window.open('/scielo.php?script=sci_nlinks&ref=2046336&pid=S1405-5546200300040000600009&lng=','','width=640,height=500,resizable=yes,scrollbars=1,menubar=yes,');">Links</a>&#160;]<!-- end-ref --><!-- ref --><p align="justify"><font face="verdana" size="2">10. <b>T. Solorio. </b>Using unlabeled data to improve classifier accuracy. Master's thesis, Computer Science Department, Instituto Nacional de Astrof&iacute;sica, &Oacute;ptica y Electr&oacute;nica, 2002.</font>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[&#160;<a href="javascript:void(0);" onclick="javascript: window.open('/scielo.php?script=sci_nlinks&ref=2046337&pid=S1405-5546200300040000600010&lng=','','width=640,height=500,resizable=yes,scrollbars=1,menubar=yes,');">Links</a>&#160;]<!-- end-ref --><!-- ref --><p align="justify"><font face="verdana" size="2">11. <b>T. Solorio </b>and<b> O. Fuentes. </b>Improving classifier accuracy using unlabeled data. International Conference on Artificial Intelligence and Applications, 2001.</font>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[&#160;<a href="javascript:void(0);" onclick="javascript: window.open('/scielo.php?script=sci_nlinks&ref=2046338&pid=S1405-5546200300040000600011&lng=','','width=640,height=500,resizable=yes,scrollbars=1,menubar=yes,');">Links</a>&#160;]<!-- end-ref --><!-- ref --><p align="justify"><font face="verdana" size="2">12. <b>M. A. Turk </b>and<b> A. P. Pentland. </b>Face recognition using eigenfaces. Proc. of IEEE Conf. on Computer Vision and Pattern Recognition, pages 586&#150;591, 1991.</font>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;[&#160;<a href="javascript:void(0);" onclick="javascript: window.open('/scielo.php?script=sci_nlinks&ref=2046339&pid=S1405-5546200300040000600012&lng=','','width=640,height=500,resizable=yes,scrollbars=1,menubar=yes,');">Links</a>&#160;]<!-- end-ref --> ]]></body><back>
<ref-list>
<ref id="B1">
<label>1</label><nlm-citation citation-type="book">
<person-group person-group-type="author">
<name>
<surname><![CDATA[Bishop]]></surname>
<given-names><![CDATA[C. M.]]></given-names>
</name>
</person-group>
<source><![CDATA[Neural Networks for Pattern Recognition]]></source>
<year>1996</year>
<publisher-loc><![CDATA[Oxford ]]></publisher-loc>
<publisher-name><![CDATA[Oxford Universiy Press]]></publisher-name>
</nlm-citation>
</ref>
<ref id="B2">
<label>2</label><nlm-citation citation-type="confpro">
<person-group person-group-type="author">
<name>
<surname><![CDATA[Blum]]></surname>
<given-names><![CDATA[A]]></given-names>
</name>
<name>
<surname><![CDATA[Mitchell]]></surname>
<given-names><![CDATA[T]]></given-names>
</name>
</person-group>
<source><![CDATA[Learning to classify text from labeled and unlabeled documents]]></source>
<year></year>
<conf-name><![CDATA[ Conference on Computational Learning Theory]]></conf-name>
<conf-date>1998</conf-date>
<conf-loc> </conf-loc>
</nlm-citation>
</ref>
<ref id="B3">
<label>3</label><nlm-citation citation-type="book">
<person-group person-group-type="author">
<name>
<surname><![CDATA[Dietterich]]></surname>
<given-names><![CDATA[T. G.]]></given-names>
</name>
</person-group>
<source><![CDATA[Machine learning research: Four current directions]]></source>
<year>1997</year>
<publisher-name><![CDATA[The AI Magazine]]></publisher-name>
</nlm-citation>
</ref>
<ref id="B4">
<label>4</label><nlm-citation citation-type="book">
<person-group person-group-type="author">
<name>
<surname><![CDATA[Fausett]]></surname>
<given-names><![CDATA[L]]></given-names>
</name>
</person-group>
<source><![CDATA[Fundamentals of Neural Networks: Architectures, Algorithms and Aplications]]></source>
<year>1994</year>
<publisher-name><![CDATA[Prentice- Hall]]></publisher-name>
</nlm-citation>
</ref>
<ref id="B5">
<label>5</label><nlm-citation citation-type="book">
<person-group person-group-type="author">
<name>
<surname><![CDATA[Freund]]></surname>
<given-names><![CDATA[Y]]></given-names>
</name>
<name>
<surname><![CDATA[Seung]]></surname>
<given-names><![CDATA[H]]></given-names>
</name>
<name>
<surname><![CDATA[Shamir]]></surname>
<given-names><![CDATA[E]]></given-names>
</name>
<name>
<surname><![CDATA[Tishby]]></surname>
<given-names><![CDATA[N]]></given-names>
</name>
</person-group>
<source><![CDATA[Selective sampling using the query by committee algorithm]]></source>
<year>1997</year>
<publisher-name><![CDATA[Machine Learning]]></publisher-name>
</nlm-citation>
</ref>
<ref id="B6">
<label>6</label><nlm-citation citation-type="book">
<person-group person-group-type="author">
<name>
<surname><![CDATA[Graham]]></surname>
<given-names><![CDATA[D]]></given-names>
</name>
<name>
<surname><![CDATA[Allinson]]></surname>
<given-names><![CDATA[N]]></given-names>
</name>
</person-group>
<article-title xml:lang="en"><![CDATA[Character-rising Virtual Eigensignatures for Face Recognition]]></article-title>
<source><![CDATA[Face Recognition: From Theory to Applications]]></source>
<year>1998</year>
<publisher-name><![CDATA[Springer-Verlag]]></publisher-name>
</nlm-citation>
</ref>
<ref id="B7">
<label>7</label><nlm-citation citation-type="confpro">
<person-group person-group-type="author">
<name>
<surname><![CDATA[McCallum]]></surname>
<given-names><![CDATA[A]]></given-names>
</name>
<name>
<surname><![CDATA[Nigam]]></surname>
<given-names><![CDATA[K]]></given-names>
</name>
</person-group>
<article-title xml:lang="en"><![CDATA[Employing EM in pool-based active learning for text classification]]></article-title>
<source><![CDATA[]]></source>
<year></year>
<conf-name><![CDATA[ Proceeding of the 15th International Conference on Machine Learning]]></conf-name>
<conf-date>1998</conf-date>
<conf-loc> </conf-loc>
</nlm-citation>
</ref>
<ref id="B8">
<label>8</label><nlm-citation citation-type="book">
<person-group person-group-type="author">
<name>
<surname><![CDATA[Mitchell]]></surname>
<given-names><![CDATA[T. M.]]></given-names>
</name>
</person-group>
<source><![CDATA[Machine Learning]]></source>
<year>1997</year>
<publisher-name><![CDATA[McGraw-Hill]]></publisher-name>
</nlm-citation>
</ref>
<ref id="B9">
<label>9</label><nlm-citation citation-type="book">
<person-group person-group-type="author">
<name>
<surname><![CDATA[Nigam]]></surname>
<given-names><![CDATA[K]]></given-names>
</name>
<name>
<surname><![CDATA[McCallum]]></surname>
<given-names><![CDATA[A]]></given-names>
</name>
<name>
<surname><![CDATA[Thrun]]></surname>
<given-names><![CDATA[S]]></given-names>
</name>
<name>
<surname><![CDATA[Mitchell]]></surname>
<given-names><![CDATA[T]]></given-names>
</name>
</person-group>
<source><![CDATA[Learning to classify text from labeled and unlabeled documents]]></source>
<year>1999</year>
<publisher-name><![CDATA[Machine Learning]]></publisher-name>
</nlm-citation>
</ref>
<ref id="B10">
<label>10</label><nlm-citation citation-type="">
<person-group person-group-type="author">
<name>
<surname><![CDATA[Solorio]]></surname>
<given-names><![CDATA[T]]></given-names>
</name>
</person-group>
<source><![CDATA[Using unlabeled data to improve classifier accuracy]]></source>
<year></year>
</nlm-citation>
</ref>
<ref id="B11">
<label>11</label><nlm-citation citation-type="confpro">
<person-group person-group-type="author">
<name>
<surname><![CDATA[Solorio]]></surname>
<given-names><![CDATA[T]]></given-names>
</name>
<name>
<surname><![CDATA[Fuentes]]></surname>
<given-names><![CDATA[O]]></given-names>
</name>
</person-group>
<source><![CDATA[Improving classifier accuracy using unlabeled data]]></source>
<year></year>
<conf-name><![CDATA[ International Conference on Artificial Intelligence and Applications]]></conf-name>
<conf-date>2001</conf-date>
<conf-loc> </conf-loc>
</nlm-citation>
</ref>
<ref id="B12">
<label>12</label><nlm-citation citation-type="confpro">
<person-group person-group-type="author">
<name>
<surname><![CDATA[Turk]]></surname>
<given-names><![CDATA[M. A.]]></given-names>
</name>
<name>
<surname><![CDATA[Pentland]]></surname>
<given-names><![CDATA[A. P.]]></given-names>
</name>
</person-group>
<source><![CDATA[Face recognition using eigenfaces]]></source>
<year>1991</year>
<conf-name><![CDATA[ Proc. of IEEE Conf. on Computer Vision and Pattern Recognition]]></conf-name>
<conf-loc> </conf-loc>
<page-range>586-591</page-range></nlm-citation>
</ref>
</ref-list>
</back>
</article>
