1. DSpace at Kasdi Merbah University Ouargla
  2. 3. Mémoires de Master
  3. 2. Faculté des nouvelles technologies de l’information et de la communication
  4. Département d'Electronique et des Télécommunications - Master
Please use this identifier to cite or link to this item: https://dspace.univ-ouargla.dz/jspui/handle/123456789/21928
Full metadata record
DC FieldValueLanguage
dc.contributor.authorAfafe. Lahreche, Abdelhakim. Cheriet-
dc.date.accessioned2019-11-11T07:46:18Z-
dc.date.available2019-11-11T07:46:18Z-
dc.date.issued2019-11-11-
dc.identifier.urihttp://dspace.univ-ouargla.dz/jspui/handle/123456789/21928-
dc.description.abstractThe main objectives of our work is using the e ectiveness of the Deep learning methods for Arabic manuscript recognition. In this work, we used Convolution Neural Networks where we train and test them using Tensor ow library that support machine learning in Python in order to obtain the required results.en_US
dc.description.sponsorshipUNIVERSITY KASDI-MERBAH OUARGLA Faculty of New Technologies of Information and Communication Department of Computer Science and Information Technologiesen_US
dc.language.isoenen_US
dc.subjectHandwriting, deep learning, Neural Networks, Convolution Neural Net- works. R esum e Les principaux objectifs de notre travail est l'utilisation de l' e cacit e du Deep learning m ethodes pour la reconnaissance des manuscrits en arabe. Dans ce travail, jai utilis e des r eseau neuronal convolutif o u nous les avons form es et test es a laide de biblioth eques Tensor ow qui prend en charge lapprentissage automatique en Python a n dobtenir les r esultats requis. Mots-cl es: ecriture manuscrite, apprentissage en profondeur, r eseaux de neurones, r eseaux de neurones de convolution. 1en_US
dc.titleDeep learning for Arabic letters recognitionen_US
Appears in Collections:Département d'Electronique et des Télécommunications - Master

Files in This Item:
File Description SizeFormat 
Deep learning for Arabic letters recognition.pdf454,74 kBAdobe PDFView/Open
Show simple item record


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.