Automatic recognition of Arabic alphabets sign language using deep learning

<span>Technological advancements are helping people with special needs overcome many communications’ obstacles. Deep learning and computer vision models innovative leaps nowadays in facilitating unprecedented tasks human interactions. The Arabic language is always a rich research area. In this paper, different deep were applied to test the accuracy efficiency obtained automatic sign recognition. we provide novel framework for detection of language, based on transfer popular image processing. Specifically, by training AlexNet, VGGNet GoogleNet/Inception models, along testing shallow approaches support vector machine (SVM) nearest neighbors algorithms as baselines. As result, propose approach recognition alphabets architecture which outperformed other trained models. proposed model set present promising results recognizing an score 97%. suggested tested against recent fully-labeled dataset images. contains 54,049 images, considered first large comprehensive real furthest know.</span>