Isharah Continuous Sign Language Recognition and Translation Dataset

The first large-scale continuous Saudi Sign Language (SSL) dataset

🧾 About

Isharah is a large-scale dataset for Continuous Saudi Sign Language (SSL) recognition and translation. It features over 30,000 video samples signed by deaf and hearing-impaired individuals using smartphones in varied settings.

The dataset supports both Continuous Sign Language Recognition (CSLR) and Sign Language Translation (SLT), and includes sentence-level gloss annotations and corresponding Arabic translations. Three benchmark subsets are included: Isharah-500, Isharah-1000, and Isharah-2000.

🔗 Download

Please register your details and accept the usage agreement to access Isharah-500, Isharah-1000, and pose data.

Access the Dataset

Isharah-2000: coming soon.

🏆 Related Challenges

Isharah has been used in the SignEval challenges for benchmarking continuous sign language recognition and related sign language understanding tasks.

SignEval 2026

Multi-Modal Sign Language Recognition and Translation Challenge

SignEval 2025

Multi-Modal Sign Language Recognition Challenge at ICCV 2025

📄 Citation

If you use Isharah in your work, please cite:

@ARTICLE{11397217,
  author={Alyami, Sarah and Luqman, Hamzah and Al-Azani, Sadam and Alowaifeer, Maad and Alharbi, Yazeed and Alonaizan, Yaser},
  journal={IEEE Transactions on Multimedia},
  title={Isharah: A Large-Scale Multi-Scene Dataset for Continuous Sign Language Recognition},
  year={2026},
  volume={},
  number={},
  pages={1-9},
  keywords={Sign Language Recognition;Continuous Sign Language Recognition;Sign Language Translation;Arabic Sign Language;Sign Language Dataset},
  doi={10.1109/TMM.2026.3664959}}