Researchers have introduced SyriSign, a new dataset designed to facilitate the translation of Arabic text into Syrian Arabic Sign Language (SyArSL). This dataset contains 1500 video samples of 150 unique lexical signs, aiming to bridge the communication gap for the deaf and hard-of-hearing community in Syria who often lack access to information presented in spoken or written Arabic. The dataset was evaluated using deep learning models like MotionCLIP, T2M-GPT, and SignCLIP, though the limited size of SyriSign currently restricts generalization performance. AI
IMPACT This dataset could improve accessibility of information for deaf communities in Arabic-speaking regions.
RANK_REASON The item describes a new academic dataset and its evaluation, fitting the research bucket. [lever_c_demoted from research: ic=1 ai=1.0]
- Arabic
- deaf and hard of hearing people
- Khloud Al Jallad
- MotionCLIP
- SignCLIP
- Syria
- Syrian Arabic Sign Language
- SyriSign
- T2M-GPT
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
