Presentation of simultaneous speech translation
| Název práce v češtině: | Prezentování výstupu simultánního strojového překladu mluvené řeči |
|---|---|
| Název v anglickém jazyce: | Presentation of simultaneous speech translation |
| Klíčová slova: | simultánní překlad mluvené řeči|automatické titulkování|uživatelské vyhodnocování |
| Klíčová slova anglicky: | simultaneous speech translation|automatic subtitling|user evaluation |
| Akademický rok vypsání: | 2021/2022 |
| Typ práce: | diplomová 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: | 26.05.2021 |
| Datum zadání: | 03.06.2021 |
| Datum potvrzení stud. oddělením: | 22.06.2021 |
| Datum a čas obhajoby: | 08.09.2021 09:00 |
| Datum odevzdání elektronické podoby: | 22.07.2021 |
| Datum odevzdání tištěné podoby: | 22.07.2021 |
| Datum proběhlé obhajoby: | 08.09.2021 |
| Oponenti: | doc. Ing. Zdeněk Žabokrtský, Ph.D. |
| Zásady pro vypracování |
| The goal of the thesis is to examine methods for presenting the outputs of simultaneous speech translation systems to users.
Different from movie subtitles which are manually edited and heavily influenced by limitations of reading speed and space constraints, current speech recognition systems and speech translation systems at their best produce verbatim transcription and translation of what has been said. They lack the capabilities of reasonably simplifying and compressing the text to meet any presentation constraints. The thesis will work with the current speech translation system and examine methods of presenting their output to the user in real time, even if the systems cannot compress the content. A user interface will be proposed and tested in realistic conditions with a small group of subjects watching videos in a language they do not speak, relying on the provided automatic speech recognition, translation and presentation. The goal is to come up with an evaluation methodology and then to find the best layout from a number of presentation options. The thesis should rely on existing speech recognition and translation components, focusing purely on the presentation. Time permitting, a nice extension of the thesis would be to train machine translation systems so that they would be also compressing the text. Such a shortening system should be again empirically evaluated, including the presentation interface. |
| Seznam odborné literatury |
| Dominik Macháček, Ondřej Bojar (2020): Presenting Simultaneous Translation in Limited Space. In: Proceedings of the 20th Conference Information Technologies - Applications and Theory (ITAT 2020), pp. 32-37, Tomáš Horváth, Košice, Slovakia.
Alina Karakanta, Matteo Negri, and Marco Turchi. Is 42 the answer to everything in subtitling-oriented speech translation? 2020. Eunah Cho, C. Fugen, T. Hermann, K. Kilgour, Mohammed Mediani, C. Mohr, J. Niehues, Kay Rottmann, C. Saam, Sebastian Sẗuker, and A. Waibel. 2013. A real-world system for simultaneous translation of german lectures. pages 3473–3477. Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. 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. |