A Study of Convolutional Architectures for Handshape Recognition applied to Sign Language

cic.institucionOrigenInstituto de Investigación en Informáticaes
cic.isFulltexttruees
cic.isPeerReviewedtruees
cic.lugarDesarrolloInstituto de Investigación en Informáticaes
cic.versioninfo:eu-repo/semantics/publishedVersiones
dc.date.accessioned2018-11-29T17:15:10Z
dc.date.available2018-11-29T17:15:10Z
dc.identifier.urihttps://digital.cic.gba.gob.ar/handle/11746/8630
dc.titleA Study of Convolutional Architectures for Handshape Recognition applied to Sign Languageen
dc.typeDocumento de conferenciaes
dcterms.abstractConvolutional Neural Networks have been providing a performance boost in many areas in the last few years, but their performance for Handshape Recognition in the context of Sign Language Recognition has not been thoroughly studied. We evaluated several convolutional architectures in order to determine their applicability for this problem. Using the LSA16 and RWTH-PHOENIX-Weather handshape datasets, we performed experiments with the LeNet, VGG16, ResNet-34 and All Convolutional architectures, as well as Inception with normal training and via transfer learning, and compared them to the state of the art in these datasets. We included experiments with a feedforward neural network as a baseline. We also explored various preprocessing schemes to analyze their impact on the recognition. We determined that while all models perform reasonably well on both datasets (with performance similar to hand-engineered methods), VGG16 produced the best results, closely followed by the traditional LeNet architecture. Also, pre-segmenting the hands from the background provided a big boost to accuracy.en
dcterms.creator.authorQuiroga, Facundoes
dcterms.creator.authorAntonio, Ramiroes
dcterms.creator.authorRonchetti, Francoes
dcterms.creator.authorLanzarini, Laura Cristinaes
dcterms.creator.authorRosete, Alejandroes
dcterms.extentp. 13-22es
dcterms.identifier.otherhdl:10915/63481es
dcterms.isPartOf.seriesXXIII Congreso Argentino de Ciencias de la Computación (La Plata, 2017)es
dcterms.issued2017-10
dcterms.languageIngléses
dcterms.licenseAttribution-NonCommercial-ShareAlike 4.0 International (BY-NC-SA 4.0)es
dcterms.subjectconvolutional neural networksen
dcterms.subjectsign language recognitionen
dcterms.subjecthandshape recognitionen
dcterms.subject.materiaCiencias de la Computaciónes

Archivos

Bloque original

Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
Documento_completo.pdf-PDFA.pdf
Tamaño:
913.66 KB
Formato:
Adobe Portable Document Format
Descripción:
Documento completo