Enhancing a PEMT Task with Speech Synthesis– Getting Inspiration from the Audio in the Audiovisual

Description

The multimodal aspect of audiovisual translation has been written about as both a challenge and an opportunity – it requires a translation to be in synch with the visual and audio channels while supporting meaning making for both translators and the target viewers. We set out to explore if this positive effect can extend beyond AVT and support a neural machine translation (NMT) post editing task (PEMT). We investigate if, by accessing the text to be translated also in an audio format delivered via synthetic voices can support translators in their PEMT work. This research was also inspired by the rapid development of synthetic voices and their ubiquity in everyday life.
The results are part of the project “The impact of Speech Synthesis on cognitive load, productivity and quality during post-editing machine translation (PEMT)” funded by Imminent (the research arm of Translated) and carried out at the University of Vienna HAITrans Research group led by Univ.-Prof. Dragoș Ciobanu https://haitrans.univie.ac.at/research/imminent-the-impact-of-speech-synthesis-on-cognitive-load-productivity-and-quality-during-post-editing-machine-translation-pemt/

Period	15 Nov 2024
Event title	Languages & the Media: 15th International Conference and Exhibition on Language Transfer in Audiovisual Media
Event type	Conference
Location	Budapest, HungaryShow on map
Degree of Recognition	International