Deaf & Dumb (Tafahm mobile application)

Arabic Correction Model Integrated into a Real-Time ASL App

Developed a specialized Arabic grammatical and spelling correction model designed for integration into a real-time Arabic Sign Language (ASL) mobile application for the Deaf community. The system enhances communication by automatically correcting Arabic text generated from ASL translation, ensuring accuracy, clarity, and readability.

The model leverages advanced Natural Language Processing (NLP) techniques, including sequence-to-sequence architectures and transformer-based approaches, trained on a carefully curated Arabic text corpus. It supports Modern Standard Arabic with the ability to handle common grammatical, morphological, and orthographic errors.

Key Features:

Real-time correction for ASL-to-text conversion output

Supports grammatical, morphological, and spelling adjustments

Optimized for Modern Standard Arabic with potential dialect handling

Seamlessly integrates into mobile applications

Enhances accessibility and usability for the Deaf community

Impact:

This solution bridges the gap between ASL-generated text and formal written Arabic, improving inclusivity and ensuring effective communication for Deaf and hard-of-hearing users.