Automatic Spoken Language Translation in Real-World Settings (ASLTRW)

Workshop inside the conference Machine Translation Summit 2021

20 August 2021



Call for Papers

To date, the production of audio-video content may have exceeded that of written texts. The need to make such content available across language barriers has increased the interest in spoken language translation (SLT), opening up new opportunities for the use of speech translation applications in different settings and for different scopes, such as live translation at international conferences, automatic subtitling for video accessibility, automatic or human-in-the-loop respeaking, or as a support system for human interpreters, to name just a few. Furthermore, specific needs are emerging in terms of user profiles, e.g. people with different abilities, and user experiences, e.g. use on mobile devices.

Against this backdrop, the Spoken Language Translation in Real-World Settings workshop aims to bring together researchers in the areas of computer science, translation, and interpreting, as well as users of SLT applications, such as international organizations, businesses, broadcasters, content media creators, to discuss the latest advances in speech translation technologies from both the perspective of the Computer Science and the Humanities, raising awareness on topics such as the challenges in evaluating current technologies in real-life scenarios, customization tools to improve performance, ethical issues, human-machine interaction, and so forth.

The online workshop will seek to answer these questions by inviting papers exploring topics including, but not limited to:

  • Latest advances in cascading and end-to-end systems
  • Use of spoken language translation system in a real-world scenario
  • Automatic subtitling and dubbing
  • Communicative inclusion through SLT
  • Automatic or semi-automatic respeaking
  • Customization of SLT systems to real-world applications
  • Computer-assisted interpreting
  • Data scarcity and underrepresented languages
  • Potential and risks of SLT
  • Ethical issues (bias, etc.)
  • SLT evaluation
  • Communicative-oriented evaluation of spoken language translation
  • Benchmark creation
  • Human-machine interaction
  • Hybrid systems for spoken language translation
  • Cross-lingual NER/summarization/retrieval from speech
  • Automatic generation of meeting minuting

Program & Invited Speakers

09:00 - 09:10 Opening: Claudio Fantinuoli & Marco Turchi
09:10 - 09:55

Invited Talk

Juan Pino, Facebook AI (bio)

End-to-end Speech Translation at Facebook

09:55 - 10:15 Paper: Operating a Complex SLT System with Speakers and Human Interpreters,

Ondřej Bojar, Vojtěch Srdečný, Rishu Kumar, Otakar Smrž, Felix Schneider, Barry Haddow, Phil Williams, Chiara Canton

10:15 - 10:35

Paper: Simultaneous Speech Translation for Live Subtitling: from Delay to Display

Alina Karakanta, Sara Papi, Matteo Negri, Marco Turchi

10:35 - 10:55

Paper: Post-Editing Job Profiles for Subtitlers

Anke Tardel, Silvia Hansen-Schirra, Jean Nitzke

10:55 - 11:15

Paper: Technology-Augmented Multilingual Communication Models

Francesco Saina

11:15 - 12:00

Invited Talk

Bart Defrancq, Ghent University (bio)

Will it take another 19 years? Cognitive Ergonomics of Computer-Assisted Interpreting (CAI)

13:00 - 13:45

Invited talk

Silvia Hansen-Schirra, Mainz University (bio)

CompAsS - Computer-Assisted Subtitling

13:45 - 14:05 Paper: Seed Words Based Data Selection for Language Model Adaptation

Roberto Gretter, Marco Matassoni, Daniele Falavigna

14:05 - 14:50

Invited talk

Marcello Federico, AMAZON AI (bio)

Recent Efforts on Automatic Dubbing

14:50 - max. 16:00


Closing remarks: Claudio Fantinuoli & Marco Turchi


Scientific committee

Name Affiliation
Nguyen Bach Alibaba US
Laurent Besacier University of Grenoble
Dragos Ciobanu University of Vienna
Jorge Civera Universitat Politècnica de València
Marta R. Costa-jussà Universitat Politècnica de Catalunya
Bart Defrancq Ghent University
Marco Gaido Fondazione Bruno Kessler
Hirofumi Inaguma University of Kyoto
Alfons Juan Universitat Politècnica de València
Alina Karakanta Fondazione Bruno Kessler
Evgeny Matusov AppTek
Jan Niehues University of Maastricht
Sara Papi Fondazione Bruno Kessler
Franz Pöchhacker University of Vienna
Bianca Prandi Johannes Gutenberg-Universität Mainz
Pablo Romero-Fresco Universidade de Vigo
Juan Pino Facebook
Claudio Russello UNINT - Rome
Matthias Sperber Apple
Sebastian Stueker Karlsruhe Institute of Technology
Shinji Watanabe Johns Hopkins University


You are welcomed to contact the organizers for any question

Claudio Fantinuoli
Johannes Gutenberg-Universität Mainz / KUDO Inc.
E-Mail: fantinuoli(at)

Marco Turchi
Fondazione Bruno Kessler
E-Mail: turchi(at)

Declaration of financial support: The author(s) received financial support for the organization of this workshop from KUDO Inc. The support covered the conference fees of the invited speakers.