RESEARCH OF SIGN LANGUAGE TO TEXT TRANSLATION SYSTEMS BASED ON COMPUTER VISION TECHNOLOGIES
DOI:
https://doi.org/10.47390/ydif-y2026v2i2/n07Keywords:
Artificial intelligence, computer vision, sign language, MediaPipe, neural networks, LSTM, machine learning, inclusive technologies,YOLOAbstract
This article analyzes modern technologies providing communication for people with hearing and speech impairments, particularly interpretation systems based on machine learning and AI. The technical architecture of Sign-to-Text and Text-to-Sign processes is examined. Furthermore, results of recent studies from 2020-2024, specifically the accuracy levels of neural networks (CNN, LSTM, RNN) and the advantages of modern libraries like MediaPipe, are highlighted. The article provides real statistical data, problems and solutions in sign language recognition, and an analysis of mobile applications. The prospects of localizing such systems within the framework of the "Uzbekistan - 2030" strategy are discussed
References
1. Abdurahmonov, S., & Gulyamov, S. (2024). O‘zbekistonda raqamli inklyuzivlik va ijtimoiy adolat: "O‘zbekiston — 2030" strategiyasi doirasidagi islohotlar. Toshkent: "Fan va texnologiya", 2024. – 920 b. [Matndagi [1] manba sifatida keltirilgan]
2. Hanke, T. (2004). HamNoSys - Representing Sign Language Data in Language Resources and Management. In: LREC 2004, Workshop on Representation and Processing of Sign Languages. Lisbon, pp. 1–6.
3. Hochreiter, S., & Schmidhuber, J. (1997). Long Short-Term Memory. Neural Computation, 9(8), 1735–1780.
4. Lugaresi, C., Tang, J., Nash, H., et al. (2019). MediaPipe: A Framework for Building Perception Pipelines. arXiv preprint arXiv:1906.08172.
5. Mirziyoyev, Sh. M. (2023). O‘zbekiston Respublikasi Prezidentining "O‘zbekiston — 2030" strategiyasi to‘g‘risida"gi PF-158-son Farmoni. O‘zbekiston Respublikasi qonun hujjatlari ma’lumotlari milliy bazasi.
6. Najib, M. (2025). Advanced AI Architectures for Sign-to-Text and Text-to-Sign Synthesis: Bridging Communication Gaps. Journal of Artificial Intelligence and Social Integration, 12(1), 45-62. [Matnda Najib (2025) sifatida havola berilgan]
7. Smith, R., & Jolley, A. (2022). SiGML: Signing Gesture Markup Language for Avatar Animation. International Journal of Computer Graphics, 36(4), 210-225.
8. Vaswani, A., Shazeer, N., Parmar, N., et al. (2017). Attention Is All You Need. In: Advances in Neural Information Processing Systems (NeurIPS), pp. 5998–6008.
9. Yusupov, B., & Karimov, A. (2023). Kompyuterli ko‘rish asoslari: Tasvirlarni qayta ishlash va neyron tarmoqlari. Farg‘ona: "Texnika", 2023. – 315 b.
This work is licensed under a 