Scaling up ConvAtt for Sign Language Recognition

Autores
Ríos, Gastón Gustavo; Dal Bianco, Pedro Alejandro; Ronchetti, Franco; Quiroga, Facundo Manuel; Ponte Ahón, Santiago Andrés; Stanchi, Oscar Agustín; Hasperué, Waldo
Año de publicación
2024
Idioma
inglés
Tipo de recurso
documento de conferencia
Estado
versión publicada
Descripción
Sign language is crucial for communication within the deaf community, making Sign Language Recognition (SLR) essential for bridging the gap between signers and non-signers. However, SLR models often face challenges due to limited data availability and quality. This paper investigates various data augmentation and regularization techniques to enhance the performance of a lightweight SLR model. We focus on recognizing signs from the French Belgian Sign Language using a novel model architecture that integrates convolutional, channel attention, and selfattention layers. Our experiments demonstrate the effectiveness of these techniques, achieving a top-1 accuracy of 49.99% and a top-10 accuracy of 83.19% across 600 distinct signs.
Red de Universidades con Carreras en Informática
Materia
Ciencias Informáticas
Handshape Recognition
Unbalanced Data
Limited Data
Sign Language
Human Motion Prediction
Nivel de accesibilidad
acceso abierto
Condiciones de uso
http://creativecommons.org/licenses/by-nc-sa/4.0/
Repositorio
SEDICI (UNLP)
Institución
Universidad Nacional de La Plata
OAI Identificador
oai:sedici.unlp.edu.ar:10915/176284

id SEDICI_0cdb92663a09dc8e8bac197c56b13fdc
oai_identifier_str oai:sedici.unlp.edu.ar:10915/176284
network_acronym_str SEDICI
repository_id_str 1329
network_name_str SEDICI (UNLP)
spelling Scaling up ConvAtt for Sign Language RecognitionRíos, Gastón GustavoDal Bianco, Pedro AlejandroRonchetti, FrancoQuiroga, Facundo ManuelPonte Ahón, Santiago AndrésStanchi, Oscar AgustínHasperué, WaldoCiencias InformáticasHandshape RecognitionUnbalanced DataLimited DataSign LanguageHuman Motion PredictionSign language is crucial for communication within the deaf community, making Sign Language Recognition (SLR) essential for bridging the gap between signers and non-signers. However, SLR models often face challenges due to limited data availability and quality. This paper investigates various data augmentation and regularization techniques to enhance the performance of a lightweight SLR model. We focus on recognizing signs from the French Belgian Sign Language using a novel model architecture that integrates convolutional, channel attention, and selfattention layers. Our experiments demonstrate the effectiveness of these techniques, achieving a top-1 accuracy of 49.99% and a top-10 accuracy of 83.19% across 600 distinct signs.Red de Universidades con Carreras en Informática2024-10info:eu-repo/semantics/conferenceObjectinfo:eu-repo/semantics/publishedVersionObjeto de conferenciahttp://purl.org/coar/resource_type/c_5794info:ar-repo/semantics/documentoDeConferenciaapplication/pdf145-154http://sedici.unlp.edu.ar/handle/10915/176284enginfo:eu-repo/semantics/altIdentifier/isbn/978-950-34-2428-5info:eu-repo/semantics/reference/hdl/10915/172755info:eu-repo/semantics/openAccesshttp://creativecommons.org/licenses/by-nc-sa/4.0/Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)reponame:SEDICI (UNLP)instname:Universidad Nacional de La Platainstacron:UNLP2025-09-29T11:47:28Zoai:sedici.unlp.edu.ar:10915/176284Institucionalhttp://sedici.unlp.edu.ar/Universidad públicaNo correspondehttp://sedici.unlp.edu.ar/oai/snrdalira@sedici.unlp.edu.arArgentinaNo correspondeNo correspondeNo correspondeopendoar:13292025-09-29 11:47:28.773SEDICI (UNLP) - Universidad Nacional de La Platafalse
dc.title.none.fl_str_mv Scaling up ConvAtt for Sign Language Recognition
title Scaling up ConvAtt for Sign Language Recognition
spellingShingle Scaling up ConvAtt for Sign Language Recognition
Ríos, Gastón Gustavo
Ciencias Informáticas
Handshape Recognition
Unbalanced Data
Limited Data
Sign Language
Human Motion Prediction
title_short Scaling up ConvAtt for Sign Language Recognition
title_full Scaling up ConvAtt for Sign Language Recognition
title_fullStr Scaling up ConvAtt for Sign Language Recognition
title_full_unstemmed Scaling up ConvAtt for Sign Language Recognition
title_sort Scaling up ConvAtt for Sign Language Recognition
dc.creator.none.fl_str_mv Ríos, Gastón Gustavo
Dal Bianco, Pedro Alejandro
Ronchetti, Franco
Quiroga, Facundo Manuel
Ponte Ahón, Santiago Andrés
Stanchi, Oscar Agustín
Hasperué, Waldo
author Ríos, Gastón Gustavo
author_facet Ríos, Gastón Gustavo
Dal Bianco, Pedro Alejandro
Ronchetti, Franco
Quiroga, Facundo Manuel
Ponte Ahón, Santiago Andrés
Stanchi, Oscar Agustín
Hasperué, Waldo
author_role author
author2 Dal Bianco, Pedro Alejandro
Ronchetti, Franco
Quiroga, Facundo Manuel
Ponte Ahón, Santiago Andrés
Stanchi, Oscar Agustín
Hasperué, Waldo
author2_role author
author
author
author
author
author
dc.subject.none.fl_str_mv Ciencias Informáticas
Handshape Recognition
Unbalanced Data
Limited Data
Sign Language
Human Motion Prediction
topic Ciencias Informáticas
Handshape Recognition
Unbalanced Data
Limited Data
Sign Language
Human Motion Prediction
dc.description.none.fl_txt_mv Sign language is crucial for communication within the deaf community, making Sign Language Recognition (SLR) essential for bridging the gap between signers and non-signers. However, SLR models often face challenges due to limited data availability and quality. This paper investigates various data augmentation and regularization techniques to enhance the performance of a lightweight SLR model. We focus on recognizing signs from the French Belgian Sign Language using a novel model architecture that integrates convolutional, channel attention, and selfattention layers. Our experiments demonstrate the effectiveness of these techniques, achieving a top-1 accuracy of 49.99% and a top-10 accuracy of 83.19% across 600 distinct signs.
Red de Universidades con Carreras en Informática
description Sign language is crucial for communication within the deaf community, making Sign Language Recognition (SLR) essential for bridging the gap between signers and non-signers. However, SLR models often face challenges due to limited data availability and quality. This paper investigates various data augmentation and regularization techniques to enhance the performance of a lightweight SLR model. We focus on recognizing signs from the French Belgian Sign Language using a novel model architecture that integrates convolutional, channel attention, and selfattention layers. Our experiments demonstrate the effectiveness of these techniques, achieving a top-1 accuracy of 49.99% and a top-10 accuracy of 83.19% across 600 distinct signs.
publishDate 2024
dc.date.none.fl_str_mv 2024-10
dc.type.none.fl_str_mv info:eu-repo/semantics/conferenceObject
info:eu-repo/semantics/publishedVersion
Objeto de conferencia
http://purl.org/coar/resource_type/c_5794
info:ar-repo/semantics/documentoDeConferencia
format conferenceObject
status_str publishedVersion
dc.identifier.none.fl_str_mv http://sedici.unlp.edu.ar/handle/10915/176284
url http://sedici.unlp.edu.ar/handle/10915/176284
dc.language.none.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv info:eu-repo/semantics/altIdentifier/isbn/978-950-34-2428-5
info:eu-repo/semantics/reference/hdl/10915/172755
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
http://creativecommons.org/licenses/by-nc-sa/4.0/
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)
eu_rights_str_mv openAccess
rights_invalid_str_mv http://creativecommons.org/licenses/by-nc-sa/4.0/
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)
dc.format.none.fl_str_mv application/pdf
145-154
dc.source.none.fl_str_mv reponame:SEDICI (UNLP)
instname:Universidad Nacional de La Plata
instacron:UNLP
reponame_str SEDICI (UNLP)
collection SEDICI (UNLP)
instname_str Universidad Nacional de La Plata
instacron_str UNLP
institution UNLP
repository.name.fl_str_mv SEDICI (UNLP) - Universidad Nacional de La Plata
repository.mail.fl_str_mv alira@sedici.unlp.edu.ar
_version_ 1844616338267439104
score 13.070432