<![CDATA[People with hearing and speech disabilities (PHSD) continue to face barriers in accessing Quranic literacy education due to the dominance of auditory–verbal instructional approaches and the limited availability of adaptive digital learning environments. Although sign language recognition (SLR) technologies have advanced significantly, most existing systems are not aligned with pedagogically and theologically grounded Quranic learning frameworks. This study aims to develop and evaluate a deep learning-enhanced Kitabah application to support inclusive, adaptive, and technology-enhanced Quranic sign language education for PHSD learners. The study employed two approaches: (1) the development of a Kitabah-based mobile learning application integrating interactive visual–motor learning features, and (2) the implementation of a deep learning-based SLR model using ResNet-18 with transfer learning for static Hijaiyyah gesture recognition. The mobile application was evaluated through black-box testing and the System Usability Scale (SUS), while the SLR model was assessed using accuracy, precision, recall, and F1-score metrics. Results showed that all application functionalities operated successfully, with the application achieving a SUS score of 78.06, indicating good usability and accessibility. The SLR model achieved 98% classification accuracy across 31 Hijaiyyah sign classes, demonstrating strong recognition performance. These findings indicate that integrating the Kitabah method with deep learning and mobile learning technology can support progressive, inclusive, and adaptive Quranic literacy education through AI-assisted and learner-centered educational experiences for PHSD learners.]]>
Copyrights © 2026
