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

Convolutional 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 determi...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Quiroga, Facundo, Antonio, Ramiro, Ronchetti, Franco, Lanzarini, Laura Cristina, Rosete, Alejandro
Formato: Objeto de conferencia
Lenguaje:Inglés
Publicado: 2017
Materias:
Ciencias Informáticas
convolutional neural networks
sign language recognition
handshape recognition.
Acceso en línea:http://sedici.unlp.edu.ar/handle/10915/63481
Aporte de:
SEDICI (UNLP) de Universidad Nacional de La Plata
id I19-R120-10915-63481
record_format dspace
institution Universidad Nacional de La Plata
institution_str I-19
repository_str R-120
collection SEDICI (UNLP)
language Inglés
topic Ciencias Informáticas
convolutional neural networks
sign language recognition
handshape recognition.
spellingShingle Ciencias Informáticas
convolutional neural networks
sign language recognition
handshape recognition.
Quiroga, Facundo
Antonio, Ramiro
Ronchetti, Franco
Lanzarini, Laura Cristina
Rosete, Alejandro
A Study of Convolutional Architectures for Handshape Recognition applied to Sign Language
topic_facet Ciencias Informáticas
convolutional neural networks
sign language recognition
handshape recognition.
description Convolutional 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.
format Objeto de conferencia
Objeto de conferencia
author Quiroga, Facundo
Antonio, Ramiro
Ronchetti, Franco
Lanzarini, Laura Cristina
Rosete, Alejandro
author_facet Quiroga, Facundo
Antonio, Ramiro
Ronchetti, Franco
Lanzarini, Laura Cristina
Rosete, Alejandro
author_sort Quiroga, Facundo
title A Study of Convolutional Architectures for Handshape Recognition applied to Sign Language
title_short A Study of Convolutional Architectures for Handshape Recognition applied to Sign Language
title_full A Study of Convolutional Architectures for Handshape Recognition applied to Sign Language
title_fullStr A Study of Convolutional Architectures for Handshape Recognition applied to Sign Language
title_full_unstemmed A Study of Convolutional Architectures for Handshape Recognition applied to Sign Language
title_sort study of convolutional architectures for handshape recognition applied to sign language
publishDate 2017
url http://sedici.unlp.edu.ar/handle/10915/63481
work_keys_str_mv AT quirogafacundo astudyofconvolutionalarchitecturesforhandshaperecognitionappliedtosignlanguage
AT antonioramiro astudyofconvolutionalarchitecturesforhandshaperecognitionappliedtosignlanguage
AT ronchettifranco astudyofconvolutionalarchitecturesforhandshaperecognitionappliedtosignlanguage
AT lanzarinilauracristina astudyofconvolutionalarchitecturesforhandshaperecognitionappliedtosignlanguage
AT rosetealejandro astudyofconvolutionalarchitecturesforhandshaperecognitionappliedtosignlanguage
AT quirogafacundo studyofconvolutionalarchitecturesforhandshaperecognitionappliedtosignlanguage
AT antonioramiro studyofconvolutionalarchitecturesforhandshaperecognitionappliedtosignlanguage
AT ronchettifranco studyofconvolutionalarchitecturesforhandshaperecognitionappliedtosignlanguage
AT lanzarinilauracristina studyofconvolutionalarchitecturesforhandshaperecognitionappliedtosignlanguage
AT rosetealejandro studyofconvolutionalarchitecturesforhandshaperecognitionappliedtosignlanguage
bdutipo_str Repositorios
_version_ 1764820480837025792

Ejemplares similares