Towards the Primary Platform for
Language Technologies in Europe

ELITR

The fourth annual ELG conference

META-FORUM 2022

Joining the European Language Grid:

Together Towards Digital Language Equality


Brussels, Belgium

Hybrid conference

Atomium Brussels
© Adobe Stock – Sergij Figurnyi

Project expo

Project Profile

Project abbreviation: ELITR

Project name: European Live Translator

Project coordinator: Ondřej Bojar

Project consortium:

  • Charles University
  • University of Edinburgh
  • Karlsruhe Institute of Technology
  • PerVoice
  • alfatraining

Funding: Horizon 2020 Research and Innovation Programme under Grant Agreement No 825460

Project duration: January 1st, 2019 to March 31st, 2022

Main key words: Machine translation, Multilingualism, Language diversity, Automated translation, Natural language processing, Service oriented architectures

Background of the research topic: ELITR builds upon the experience of its partners in machine translation and speech processing and advances the state of the art by addressing the following challenges:

  • Highly multi-lingual translation.
  • Translation of spoken genre, live subtitling.
  • Preservation of cross-sentence context, in written documents or spoken sessions.

Additionally, we are laying foundations for automatic meeting summarization (“minuting”), to provide the NLP community with one possible clear and concise definition and starting shared tasks on the topic, see AutoMin 2021.

Goal of the project: Advancement and application of ASR and SLT in Face-to-face, Remote and Hybrid conferences, Robust and efficient systems for machine translation with many source and/or target languages, Coherent translation of larger documents and dialogues, Automatic minuting, i.e. summarizing meetings in the form of structured minutes from speech

Project abstract: ELITR, an EU-funded research and innovation project pushes the frontiers of simultaneous speech translation and document-level translation.
Centered around use cases of two user partners, the Supreme Audit Office of the Czech Republic and alfatraining, ELITR is bringing to life:

  • Live subtitling in up to 43 languages, with the major test event in April 2021: the EUROSAI 2021 congress of supreme audit institutions,
  • Speech translation in alfaview remote conferencing platform.

High quality natural language processing (NLP) is critically needed in the diverse European landscape.
ELITR is co-ordinated by Charles University, with the University of Edinburgh and Karlsruhe Institute of Technology as two research partners, PerVoice as the integration partner and alfatraining as project-funded user partner. The Supreme Audit Office is contributing from its own funds.

Publications:

  • Bojar Ondřej, Macháček Dominik, Sagar Sangeet, Smrž Otakar, Kratochvíl Jonáš, Polák Peter, Ansari Ebrahim, Mahmoudi Mohammad, Kumar Rishu, Franceschini Dario, Canton Chiara, Simonini Ivan, Nguyen Thai-Son, Schneider Felix , Stüker Sebastian, Waibel Alex, Haddow Barry, Sennrich Rico, Williams Philip: ELITR Multilingual Live Subtitling: Demo and Strategy. In: Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations, Copyright © Association for Computational Linguistics (ACL), Stroudsburg, PA, USA, ISBN 978-1-954085-05-3, pp. 271-277, 2021
  • Bojar Ondřej, Srdečný Vojtěch, Kumar Rishu, Smrž Otakar, Schneider Felix, Haddow Barry, Williams Phil, Canton Chiara: Operating a Complex SLT System with Speakers and Human Interpreters. In: Proceedings of Machine Translation Summit XVIII 1st Workshop on Automatic Spoken Language Translation in Real-World Settings, Copyright © Association for Machine Translation in the Americas, Stroudsburg, PA, USA, pp. 23-34, 2021
  • Ghosal Tirthankar, Bojar Ondřej, Singh Muskaan, Nedoluzhko Anja: Overview of the First Shared Task on Automatic Minuting (AutoMin) at Interspeech 2021. In print.