KArSL is a multi-​modality dataset

All KArSL signs are available in three modalities: RGB,  depth, and skeleton joint points.

Citing

If you use KArSL dataset, we kindly ask you to cite KArSL: Arabic Sign Language Database paper:
@article{sidig2021karsl, 
           title={KArSL: Arabic Sign Language Database}, 
           author={Sidig, Ala Addin I and Luqman, Hamzah and Mahmoud, Sabri and Mohandes, Mohamed},
            journal={ACM Transactions on Asian and Low-Resource Language Information Processing (TALLIP)}, 
           volume={20}, 
           number={1}, 
           pages={1--19}, 
           year={2021}, 
           publisher={ACM New York, NY, USA} 
           }


Another useful paper that contains results on 502 signs of KArSL dataset:


@article{​luqman2022efficient, 
           title={
An Efficient Two-Stream Network for Isolated Sign Language Recognition Using Accumulative Video Motion }, 
           author={ Luqman, Hamzah },
          journal={IEEE Access},
          volume={10},
           pages={93785--93798},
           year={2022},
           publisher={IEEE}
           }


An Efficient Two-Stream Network for Isolated Sign Language Recognition Using Accumulative Video Motion

Dataset Download

Three sets of KArSL dataset are available for researchers:

KArSL-100

KArSL-100 dataset consists of 100 dynamic signs of KArSL dataset (from signID 0071 to 0170). Please follow the links below to download it:

To download the raw video files of this dataset, please follow this link.

KArSL-190

KArSL-190 dataset consists of 190 static and dynamic signs of KArSL dataset (from signID 0001 to 0190). Please follow the links below to download it:

To download the raw video files of this dataset, please follow this link.

KArSL-502

KArSL-502 dataset consists of 502 static and dynamic signs (whole KArSL dataset signs) (from signID 0001 to 0502). Please follow the links below to download it:

To download the raw video files of this dataset, please follow this link.

Contact Information

For any ​inquiries about the KArSL dataset, please contact Hamzah Luqman ( hluqman@kfupm.edu.sa)