Multi-Source Simultaneous Speech Translation
| Název práce v češtině: | Simultánní překlad řeči z více zdrojů |
|---|---|
| Název v anglickém jazyce: | Multi-Source Simultaneous Speech Translation |
| Klíčová slova: | simultánní překlad řeči|překlad řeči|překlad z řeči do textu|strojový překlad|vícejazyčnost|vícezdrojovost|simultánní tlumočení|zpracování přirozeného jazyka |
| Klíčová slova anglicky: | simultaneous speech translation|speech translation|speech-to-text translation|machine translation|multilinguality|multi-sourcing|simultaneous interpreting|natural language processing |
| Akademický rok vypsání: | 2018/2019 |
| Typ práce: | disertační práce |
| Jazyk práce: | angličtina |
| Ústav: | Ústav formální a aplikované lingvistiky (32-UFAL) |
| Vedoucí / školitel: | doc. RNDr. Ondřej Bojar, Ph.D. |
| Řešitel: | skrytý - zadáno a potvrzeno stud. odd. |
| Datum přihlášení: | 17.06.2019 |
| Datum zadání: | 17.06.2019 |
| Datum potvrzení stud. oddělením: | 04.10.2019 |
| Datum odevzdání elektronické podoby: | 11.03.2024 |
| Datum odevzdání tištěné podoby: | 22.03.2024 |
| Zásady pro vypracování |
| Neural machine translation (NMT) has the capability of handling more source and/or target languages at once.
The goal of the thesis is to experimentally explore this area and propose and evaluate variations of NMT model architectures, training data layout or training methods to achieve gains in translation quality or efficiency. Depending on the results of experiments carried out in the first stage of the studies, the work may focus primarily on one of the following use cases: - Multi-target MT, where the same input is to be translated simultaneously into multiple languages. The desired savings would be primarily in terms of memory and computing resources (one model serving more target languages), at as little loss in translation quality as possible. Thanks to GPU parallelization, multiple target languages could be produced synchronously. - Multi-source MT, where the same input sentence is available in more than one language. The expected gains would be in translation quality, thanks to the reduced ambiguity of the input. The proposed architecture should gracefully handle if not all input language versions are available, including the situation where only one source language is given. |
| Seznam odborné literatury |
| Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is All you Need. In I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, editors, Advances in Neural Information Processing Systems 30, pages 6000–6010. Curran Associates, Inc., 2017.
Orhan Firat, Kyunghyun Cho, and Yoshua Bengio. Multi-way, multilingual neural machine translation with a shared attention mechanism. In Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 866–875, San Diego, California, June 2016. Association for Computational Linguistics. Goodfellow, I., Y. Bengio, and A. Courville 2016. Deep learning. Cambridge, MA, USA: MIT press. Helcl Jindřich, Libovický Jindřich, Kocmi Tom, Musil Tomáš, Cífka Ondřej, Variš Dušan, Bojar Ondřej: Neural Monkey: The Current State and Beyond. In: The 13th Conference of The Association for Machine Translation in the Americas, Vol. 1: MT Researchers’ Track, Copyright © The Association for Machine Translation in the Americas, Stroudsburg, PA, USA, pp. 168-176, 2018 |