Automated Arabic Algerian Sign Language Translation System Based on 3D Avatar Technology

Abstract

Aiming to reduce communication barriers between deaf individuals using Arabic Algerian Sign Language (AASL) and hearing individuals unfamiliar with this language, an automated translation system has been developed. This system uses a 3D avatar to dynamically and accurately represent signs. The methodology includes the Notation System Method (NSM), which primarily relies on encoding sign language using the Hamburg Notation System (HamNoSys), and the Gesture Animation via Motion Capture (GAMC) method for recognizing and animating gestures using intelligent motion capture models. The study provides a comprehensive analysis of the linguistic complexities of AASL and the challenges in developing technological solutions to address these issues. To enhance translation accuracy, collaboration with experts in AASL translation was conducted at each development stage. The results indicate significant improvements in translation accuracy and user engagement, highlighting the system’s potential to improve accessibility for the deaf community in various fields in Algeria, based on the differences revealed through the comparison of the two methods employed.