AUDIO ELEKTRON TARJIMA TIZIMLARI: YONDASHUVLAR, ARXITEKTURALAR VA ISTIQBOLLAR
Keywords:
Kalit so'zlar: audio tarjima, nutqdan-nutqqa tarjima, avtomatik nutqni tanish, mashina tarjimasi, nutq sintezi, neyron tarmoqlar, kam resursli tillar, turkiy tillar.Abstract
Annotatsiya: Ushbu maqolada audio elektron tarjima (nutqdan-nutqqa tarjima, Speech-to-Speech Translation) tizimlarining asosiy komponentlari — avtomatik nutqni tanish (ASR), mashina tarjimasi (MT) va nutq sintezi (TTS) — tahlil qilingan. Ushbu uch komponentni birlashtirishning ikki asosiy yondashuvi, ya'ni kaskadli va end-to-end (bir bosqichli) arxitekturalar qiyoslangan, ularning afzalliklari va kamchiliklari ko'rsatib o'tilgan. Shuningdek, so'nggi yillarda paydo bo'lgan zamonaviy neyron tarmoq asosidagi modellar (transformer arxitekturasi, o'z-o'zini nazorat qiluvchi oldindan o'qitish usullari) sharhlangan. Maqolaning alohida qismi kam resursli tillar, jumladan turkiy tillar oilasi uchun audio tarjima tizimlarini yaratishdagi o'ziga xos qiyinchiliklarga bag'ishlangan: parallel korpuslarning yetishmasligi, agglyutinativ morfologiya va dialektal o'zgaruvchanlik kabi masalalar ko'rib chiqilgan. Maqola yakunida sohaning istiqbolli tadqiqot yo'nalishlari belgilab berilgan.References
1. Vaswani A., Shazeer N., Parmar N. va boshq. Attention Is All You Need // Advances in Neural Information Processing Systems (NeurIPS). — 2017. — P. 5998–6008.
2. Radford A., Kim J.W., Xu T. va boshq. Robust Speech Recognition via Large-Scale Weak Supervision (Whisper) // Proceedings of the 40th International Conference on Machine Learning. — 2023.
3. Baevski A., Zhou Y., Mohamed A., Auli M. wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations // Advances in Neural Information Processing Systems (NeurIPS). — 2020.
4. Barrault L., Chung Y-A., Meglioli M.C. va boshq. Seamless: Multilingual Expressive and Streaming Speech Translation // Meta AI Research. — 2023.
5. Ren Y., Hu C., Tan X. va boshq. FastSpeech 2: Fast and High-Quality End-to-End Text to Speech // International Conference on Learning Representations (ICLR). — 2021.
6. Kim J., Kong J., Son J. Conditional Variational Autoencoder with Adversarial Learning for End-to-End Text-to-Speech (VITS) // Proceedings of the 38th International Conference on Machine Learning. — 2021.
7. Jia Y., Ramanovich M.T., Remez T., Pomerantz R. Translatotron 2: High-quality direct speech-to-speech translation with voice preservation // Proceedings of the 39th International Conference on Machine Learning. — 2022.
8. Conneau A., Khandelwal K., Goyal N. va boshq. Unsupervised Cross-lingual Representation Learning at Scale (XLM-R) // Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. — 2020.
9. Sennrich R., Haddow B., Birch A. Improving Neural Machine Translation Models with Monolingual Data // Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics. — 2016.
10. Ruder S. A Survey of Multilingual Neural Machine Translation Approaches for Low-Resource Languages // Journal of Artificial Intelligence Research. — 2022.
