Dataset Citation
If you use this codebase or the KArSL dataset in your research, please link the original authors.
KArSL-502
Title: KArSL: Database for Word-Level Arabic Sign Language Recognition Authors: Details to be confirmed from the official publication. Institution: King Saud University
Acknowledgements
- MediaPipe: For the robust landmark extraction framework.
- PyTorch: For the deep learning training infrastructure.