INTERNATIONAL JOURNAL OF SCIENTIFIC DEVELOPMENT AND RESEARCH International Peer Reviewed & Refereed Journals, Open Access Journal ISSN Approved Journal No: 2455-2631 | Impact factor: 8.15 | ESTD Year: 2016
open access , Peer-reviewed, and Refereed Journals, Impact factor 8.15
An intelligent Arabic Sign Language Recognition system is developed using the deep learning technique. A pre-trained lightweight deep Convolutional Neural Network architecture called MobileNet-v2 is used in this work. The network is fine-tuned using the transfer learning approach to classify thirty-two different Arabic sign language letters. Recently published benchmarked Arabic Sign Language Letter (ArSL21L) is used for the experimental analysis. The model achieved a good classification performance with an F1-score of 0.808.
Keywords:
Sign Language Recognition, Deep Learning, Arabic Sign Language Letters, Transfer Learning, Fine-tuning
Cite Article:
"Arabic Sign Language Letter Recognition using MobileNet-v2 Deep Neural Network", International Journal of Science & Engineering Development Research (www.ijsdr.org), ISSN:2455-2631, Vol.7, Issue 10, page no.292 - 296, October-2022, Available :http://www.ijsdr.org/papers/IJSDR2210052.pdf
Downloads:
000337073
Publication Details:
Published Paper ID: IJSDR2210052
Registration ID:201929
Published In: Volume 7 Issue 10, October-2022
DOI (Digital Object Identifier):
Page No: 292 - 296
Publisher: IJSDR | www.ijsdr.org
ISSN Number: 2455-2631
Facebook Twitter Instagram LinkedIn