This paper presents a comprehensive review of the evolution of sign languages, with a particular emphasis on Arabic Sign Language (ArSL) and the emerging field of Quranic Sign Language (QSL). It outlines the shift from early manual-coded systems to natural sign languages and highlights technological advancements in Sign Language Recognition (SLR) and Sign Language Production (SLP) driven by deep learning architectures such as Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM) networks, and Transformers. The review critically examines existing ArSL datasets, model performances, and production systems, drawing attention to the disparity in resources between Western and Arabic sign languages. It also explores the development of QSL systems designed to support Deaf Muslims in learning Quranic content, focusing on CNN-based gesture recognition models for Hijā’īyah letters. Despite notable progress, QSL systems remain limited to isolated gesture recognition due to dataset scarcity and theological complexity. The paper concludes by identifying key challenges in temporal modeling, signer independence, and religious accuracy, and proposes future directions to enable inclusive, AI-driven Quranic education for the Deaf community.

Arabic Sign Language, CNN, Deep Learning, LSTM, Quranic Sign Language, Sign Language Recognition