Face recognition by applying wavelet subband representation and kernel associative memory
Identifieur interne : 001634 ( Main/Exploration ); précédent : 001633; suivant : 001635Face recognition by applying wavelet subband representation and kernel associative memory
Auteurs : B. L. Zhang [Australie] ; H. Zhang ; S. S. GeSource :
- IEEE Transactions on Neural Networks [ 1045-9227 ] ; 2004.
Descripteurs français
- Pascal (Inist)
English descriptors
- KwdEn :
- Association Learning, Associative memory model, Associative processing, Associative storage, Computational complexity, Experiments, Face, Face recognition, Feature extraction, Image reconstruction, Kernel methods, Mathematical models, Neural Networks (Computer), Photic Stimulation, Recognition (Psychology), Theory, Wavelet transforms.
- MESH :
Abstract
In this paper, we propose an efficient face recognition scheme which has two features: 1) representation of face images by two-dimensional (2-D) wavelet subband coefficients and 2) recognition by a modular, personalised classification method based on kernel associative memory models. Compared to PCA projections and low resolution "thumb-nail" image representations, wavelet subband coefficients can efficiently capture substantial facial features while keeping computational complexity low. As there are usually very limited samples, we constructed an associative memory (AM) model for each person and proposed to improve the performance of AM models by kernel methods. Specifically, we first applied kernel transforms to each possible training pair of faces sample and then mapped the high-dimensional feature space back to input space. Our scheme using modular autoassociative memory for face recognition is inspired by the same motivation as using autoencoders for optical character recognition (OCR), for which the advantages has been proven. By associative memory, all the prototypical faces of one particular person are used to reconstruct themselves and the reconstruction error for a probe face image is used to decide if the probe face is from the corresponding person. We carried out extensive experiments on three standard face recognition datasets, the FERET data, the XM2VTS data, and the ORL data. Detailed comparisons with earlier published results are provided and our proposed scheme offers better recognition accuracy on all of the face datasets.
Affiliations:
Links toward previous steps (curation, corpus...)
- to stream PascalFrancis, to step Corpus: 000569
- to stream PascalFrancis, to step Curation: 000221
- to stream PascalFrancis, to step Checkpoint: 000485
- to stream Main, to step Merge: 001696
- to stream PubMed, to step Corpus: 000070
- to stream PubMed, to step Curation: 000070
- to stream PubMed, to step Checkpoint: 000070
- to stream Ncbi, to step Merge: 000019
- to stream Ncbi, to step Curation: 000019
- to stream Ncbi, to step Checkpoint: 000019
- to stream Main, to step Merge: 001482
- to stream Main, to step Curation: 001634
Le document en format XML
<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en" level="a">Face recognition by applying wavelet subband representation and kernel associative memory</title><author><name sortKey="Zhang, B L" sort="Zhang, B L" uniqKey="Zhang B" first="B. L." last="Zhang">B. L. Zhang</name><affiliation wicri:level="1"><inist:fA14 i1="01"><s1>Sch. of Comp. Sci. and Mathematics Victoria University of Technology</s1><s2>Melbourne, Vic. 3011</s2><s3>AUS</s3><sZ>1 aut.</sZ></inist:fA14><country>Australie</country><wicri:noRegion>Sch. of Comp. Sci. and Mathematics Victoria University of Technology</wicri:noRegion></affiliation></author><author><name sortKey="Zhang, H" sort="Zhang, H" uniqKey="Zhang H" first="H." last="Zhang">H. Zhang</name></author><author><name sortKey="Ge, S S" sort="Ge, S S" uniqKey="Ge S" first="S. S." last="Ge">S. S. Ge</name></author></titleStmt><publicationStmt><idno type="wicri:source">INIST</idno><idno type="inist">04-0109826</idno><date when="2004">2004</date><idno type="stanalyst">PASCAL 04-0109826 EI</idno><idno type="RBID">Pascal:04-0109826</idno><idno type="wicri:Area/PascalFrancis/Corpus">000569</idno><idno type="wicri:Area/PascalFrancis/Curation">000221</idno><idno type="wicri:Area/PascalFrancis/Checkpoint">000485</idno><idno type="wicri:doubleKey">1045-9227:2004:Zhang B:face:recognition:by</idno><idno type="wicri:Area/Main/Merge">001696</idno><idno type="wicri:source">PubMed</idno><idno type="RBID">pubmed:15387257</idno><idno type="wicri:Area/PubMed/Corpus">000070</idno><idno type="wicri:Area/PubMed/Curation">000070</idno><idno type="wicri:Area/PubMed/Checkpoint">000070</idno><idno type="wicri:Area/Ncbi/Merge">000019</idno><idno type="wicri:Area/Ncbi/Curation">000019</idno><idno type="wicri:Area/Ncbi/Checkpoint">000019</idno><idno type="wicri:doubleKey">1045-9227:2004:Zhang B:face:recognition:by</idno><idno type="wicri:Area/Main/Merge">001482</idno><idno type="wicri:Area/Main/Curation">001634</idno><idno type="wicri:Area/Main/Exploration">001634</idno></publicationStmt><sourceDesc><biblStruct><analytic><title xml:lang="en" level="a">Face recognition by applying wavelet subband representation and kernel associative memory</title><author><name sortKey="Zhang, B L" sort="Zhang, B L" uniqKey="Zhang B" first="B. L." last="Zhang">B. L. Zhang</name><affiliation wicri:level="1"><inist:fA14 i1="01"><s1>Sch. of Comp. Sci. and Mathematics Victoria University of Technology</s1><s2>Melbourne, Vic. 3011</s2><s3>AUS</s3><sZ>1 aut.</sZ></inist:fA14><country>Australie</country><wicri:noRegion>Sch. of Comp. Sci. and Mathematics Victoria University of Technology</wicri:noRegion></affiliation></author><author><name sortKey="Zhang, H" sort="Zhang, H" uniqKey="Zhang H" first="H." last="Zhang">H. Zhang</name></author><author><name sortKey="Ge, S S" sort="Ge, S S" uniqKey="Ge S" first="S. S." last="Ge">S. S. Ge</name></author></analytic><series><title level="j" type="main">IEEE Transactions on Neural Networks</title><title level="j" type="abbreviated">IEEE Trans Neural Networks</title><idno type="ISSN">1045-9227</idno><imprint><date when="2004">2004</date></imprint></series></biblStruct></sourceDesc><seriesStmt><title level="j" type="main">IEEE Transactions on Neural Networks</title><title level="j" type="abbreviated">IEEE Trans Neural Networks</title><idno type="ISSN">1045-9227</idno></seriesStmt></fileDesc><profileDesc><textClass><keywords scheme="KwdEn" xml:lang="en"><term>Association Learning</term><term>Associative memory model</term><term>Associative processing</term><term>Associative storage</term><term>Computational complexity</term><term>Experiments</term><term>Face</term><term>Face recognition</term><term>Feature extraction</term><term>Image reconstruction</term><term>Kernel methods</term><term>Mathematical models</term><term>Neural Networks (Computer)</term><term>Photic Stimulation</term><term>Recognition (Psychology)</term><term>Theory</term><term>Wavelet transforms</term></keywords><keywords scheme="MESH" xml:lang="en"><term>Association Learning</term><term>Face</term><term>Neural Networks (Computer)</term><term>Photic Stimulation</term><term>Recognition (Psychology)</term></keywords><keywords scheme="Pascal" xml:lang="fr"><term>Théorie</term><term>Extraction caractéristique</term><term>Reconstruction image</term><term>Transformation ondelette</term><term>Traitement associatif</term><term>Mémoire associative</term><term>Modèle mathématique</term><term>Complexité calcul</term><term>Reconnaissance visage</term><term>Expérience</term></keywords></textClass></profileDesc></teiHeader><front><div type="abstract" xml:lang="en">In this paper, we propose an efficient face recognition scheme which has two features: 1) representation of face images by two-dimensional (2-D) wavelet subband coefficients and 2) recognition by a modular, personalised classification method based on kernel associative memory models. Compared to PCA projections and low resolution "thumb-nail" image representations, wavelet subband coefficients can efficiently capture substantial facial features while keeping computational complexity low. As there are usually very limited samples, we constructed an associative memory (AM) model for each person and proposed to improve the performance of AM models by kernel methods. Specifically, we first applied kernel transforms to each possible training pair of faces sample and then mapped the high-dimensional feature space back to input space. Our scheme using modular autoassociative memory for face recognition is inspired by the same motivation as using autoencoders for optical character recognition (OCR), for which the advantages has been proven. By associative memory, all the prototypical faces of one particular person are used to reconstruct themselves and the reconstruction error for a probe face image is used to decide if the probe face is from the corresponding person. We carried out extensive experiments on three standard face recognition datasets, the FERET data, the XM2VTS data, and the ORL data. Detailed comparisons with earlier published results are provided and our proposed scheme offers better recognition accuracy on all of the face datasets.</div></front></TEI><affiliations><list><country><li>Australie</li></country></list><tree><noCountry><name sortKey="Ge, S S" sort="Ge, S S" uniqKey="Ge S" first="S. S." last="Ge">S. S. Ge</name><name sortKey="Zhang, H" sort="Zhang, H" uniqKey="Zhang H" first="H." last="Zhang">H. Zhang</name></noCountry><country name="Australie"><noRegion><name sortKey="Zhang, B L" sort="Zhang, B L" uniqKey="Zhang B" first="B. L." last="Zhang">B. L. Zhang</name></noRegion></country></tree></affiliations></record>Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Ticri/CIDE/explor/OcrV1/Data/Main/Exploration
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 001634 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Main/Exploration/biblio.hfd -nk 001634 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien
|wiki= Ticri/CIDE
|area= OcrV1
|flux= Main
|étape= Exploration
|type= RBID
|clé= Pascal:04-0109826
|texte= Face recognition by applying wavelet subband representation and kernel associative memory
}}
| This area was generated with Dilib version V0.6.32. |  |