Center for Augmented Interpretation

We design, prototype, and do empirical research around two main topics:

  • AI-enhanced Computer-Assisted Interpreting tools
  • Machine Interpreting (speech-to-text translation)

AI-enhanced Computer-assisted Interpreting Tools

Computer-assisted interpreting (CAI) tools support professional interpreters during several phases of their work, from preparing for a highly specialized event to seeking automated support while interpreting. We are intensively working on several aspects of CAI tools: we design and develop them, test human-machine interaction, create training material to learn how to use them.

Some of our projects:

  • Development of AI-enhanced CAI tool to support interpreters with real-time information, such as numbers, terminology and proper names.
  • Testing the ergonomics of CAI tools in the human-machine interaction
  • Designing tools to generate multilingual information to assist interpreters during the meeting preparation, for example using semantic web and machine translation
  • Empirically testing how interpreters perform when using a real-time CAI tool

Machine Interpreting

Machine Interpreting (speech-to-text and speech-to-speech translation) is the automated translation of a spoken text into written text or into speech. Different from Computer Science, we study machine interpreting from a communicative perspective, i.e. how this works in real-life communications settings. We concentrate on conference interpreting.

Some of our projects:

  • Assessing different frameworks to evaluate machine interpreting from a communicative (user-centric) perspective.
  • Co-organizing the 1st Automatic Spoken Language Translation in Real-World Settings inside the MT Summit 2021.