Arabic Letter Recognition and Pronunciation Evaluation

dc.contributor.author	رطروط, حاتم
dc.contributor.author	ياسين, دلال
dc.contributor.author	حنون, مي
dc.date.accessioned	2021-06-17T10:07:41Z
dc.date.available	2021-06-17T10:07:41Z
dc.date.issued	2021
dc.description.abstract	Arabic Speech Recognition provides an automated system for recognizing the alphabet of the Arabic language using deep learning algorithms that are capable of recognizing patterns in data on its own and compiling them, and finally extracting rules from them from long periods. The most difficult thing that children face in learning at an early age (3-5 years) is the correct pronunciation of letters and words. The project aims to examine the correct pronunciation of Arabic letters for children, and it also aims to integrate learning with play through an environment that attracts children and motivates them to learn and interact. In this project, we used a convolutional neural network to identify the Arabic alphabet. We used four different models of algorithms. The first one we built our own model, the second we used the same syntax and architecture as ‘AlexNet’, the third was ‘vgg16’ using transfer learning with tuning and without tuning, and the last we used ‘Yamnet’ using transfer learning.	en_US
dc.identifier.uri	https://hdl.handle.net/20.500.11888/15791
dc.language.iso	en	en_US
dc.title	Arabic Letter Recognition and Pronunciation Evaluation	en_US
dc.type	Graduation project	en_US

Now showing 1 - 2 of 2

Now showing 1 - 1 of 1