Arabic Speech Recognition with Deep Learning: A Review

Abstract

Automatic speech recognition is the area of research concerning the enablement of machines to accept vocal input from humans and interpreting it with the highest probability of correctness. There are several techniques to implement speech recognition models. One of the emerging techniques is using neural networks with deep learning for speech recognition. Arabic is one of the most spoken languages and least highlighted in terms of speech recognition. This paper serves as a brief review on the available studies on Arabic speech recognition. In addition, it sheds some light on the services and toolkits available for Arabic speech recognition systems’ development.