New dataset SyriSign bridges communication gap for deaf Syrians by enabling Arabic-to-Syrian Sign Language translation

SyriSign introduces the first publicly available dataset for Syrian Arabic Sign Language (SyArSL) with 1,500 video samples covering 150 unique lexical signs

Dataset addresses critical gap in low-resource sign language AI, as Arabic sign languages remain severely underrepresented compared to high-resource alternatives

Three deep learning models tested: MotionCLIP for semantic motion generation, T2M-GPT for text-conditioned synthesis, and SignCLIP for bilingual embeddings

Project aims to improve accessibility for Deaf and Hard-of-Hearing (DHH) communities in Syria who cannot access news delivered in spoken or written Arabic