ArabSign

ArabSign数据集是由沙特数据与人工智能管理局-法赫德国王石油与矿业大学人工智能联合研究中心创建的，专注于连续阿拉伯手语识别。该数据集包含9,335个样本，由6名手语者执行，总录制时长约10小时。数据集通过Kinect V2相机记录，提供颜色、深度和骨骼关节点三种信息。ArabSign数据集不仅用于研究阿拉伯手语的语言结构，还旨在开发机器翻译系统，以实现阿拉伯手语与自然语言之间的转换，解决听障人士的沟通问题。

The ArabSign dataset was created by the Joint Research Center for Artificial Intelligence jointly affiliated with the Saudi Data and Artificial Intelligence Authority and King Fahd University of Petroleum and Minerals (KFUPM), focusing on continuous Arabic Sign Language (ArSL) recognition. This dataset comprises 9,335 samples collected from 6 sign language participants, with a total recording duration of approximately 10 hours. Recorded via a Kinect V2 camera, the dataset provides three modalities: color imagery, depth maps, and skeletal joint point data. The ArabSign dataset is not only used for researching the linguistic structure of Arabic Sign Language, but also aims to develop machine translation systems that enable the conversion between Arabic Sign Language and natural languages, so as to address the communication challenges faced by hearing-impaired individuals.