O‘ZBEK IMO-ISHORA TILI UCHUN DASTLABKI MA’LUMOTLAR BAZASINI SHAKLLANTIRISH VA SUN’IY INTELLEKT ASOSIDA TANIB OLISH MODELINI ISHLAB CHIQISH
This research addresses the formation of a preliminary dataset for Uzbek Sign Language and the development of an AI-based recognition model to ensure digital inclusion of persons with hearing and speech impairments in the Republic of Uzbekistan. The relevance stems from the President’s "Digital Uzbekistan – 2030" strategy and state programs supporting persons with disabilities. The methodology employed hand landmark extraction via MediaPipe, temporal sequence processing using LSTM architecture, and comparative framework analysis. As a result, 15,000 video samples for 50 gestures were collected from 30 participants, with the model achieving 80.7% accuracy on the test set. The scientific novelty lies in creating the first open dataset prototype accounting for Uzbek sign language specifics, while practical significance is demonstrated in enabling future real-time sign language interpreter systems.
1. Mirziyoyev Sh.M. “Raqamli O‘zbekiston – 2030” strategiyasi. O‘zbekiston Respublikasi Prezidenti Farmoni PF-60-son, 2020.
2. Stoll S., et al. “Sign language recognition: A deep survey.” Expert Systems with Applications, vol. 164, 2021, pp. 113794.
3. Koller O., et al. “Deep continuous sign language recognition.” CVPR Workshops, 2015, pp. 53–61.
4. Moryossef A., et al. “How2Sign: A large-scale multimodal dataset for continuous sign language recognition.” CVPR, 2021, pp. 2735–2744.
5. Hochreiter S., Schmidhuber J. “Long Short-Term Memory.” Neural Computation, vol. 9, no. 8, 1997, pp. 1735–1780.
6. Cui R., Liu H., Zhang C. “A deep neural framework for continuous sign language recognition by iterative training.” IEEE Transactions on Multimedia, vol. 21, no. 7, 2019, pp. 1880–1891.
7. Vaswani A., et al. “Attention Is All You Need.” Advances in Neural Information Processing Systems (NeurIPS), vol. 30, 2017.
8. Lugaresi C., et al. “MediaPipe: A Framework for Building Perception Pipelines.” arXiv preprint arXiv:1906.08172, 2019.
9. Karimov A., Rajabov O. “O‘zbek imo-ishora alifbosi uchun dastlabki tasniflash modeli.” O‘zbekiston fanlar akademiyasi axborotnomasi, 2022, №3, 45–52-betlar.
10. Li D., et al. “MS-ASL: A Large-Scale Dataset for American Sign Language Recognition.” IEEE FG, 2019, pp. 1–7.
Copyright (c) 2026 «ACTA NUUz»
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
.jpg)
1.png)
