Arantza Del Pozo made an extremely relevant and useful presentation on speech recognition and Machine Translation for multilingual automatic subtitling.
Perfectly in line with the very purpose of TeTra – i.e. serving as a chance to translate research into practice – her talk surveyed brand-new technology developed within EU research projects that employs machine learning techniques for more productive subtitling:
- SAVAS: speech recognition for intralingual subtitling
- SUMAT: machine translation for interlingual subtitling
- Hbb4all: automatic live multilingual subtitling
User feedback seems to be quite varied, but it generally tends to improve after initial assessment.