A Dataset for Moroccan Sign Language (MoSL)

The MoSL (Moroccan Sign Language) dataset consists of 2,199 videos representing isolated Moroccan Sign Language (MoSL) signs. The dataset includes a wide range of basic vocabulary categories such as the alphabet, numbers, pronouns, days of the week, months, seasons, and common expressions used in daily communication. The videos were provided by the Shorouk Association for the Deaf and Hard of Hearing in Ouarzazate, Morocco. They were carefully selected based on several criteria: clarity of signing, inclusion of Arabic subtitles, accurate representation of Moroccan Sign Language, and diversity among signers, encompassing both male and female participants. Each video is labeled with its corresponding Arabic subtitle, ensuring a clear mapping between the sign and its written meaning. The dataset serves as a valuable resource for automatic sign language recognition and translation research, particularly for the Moroccan dialect of Arabic Sign Language.