Evaluating Multilingual Speech Translation under Realistic Conditions with Resegmentation and Terminology
Elizabeth Salesky, Kareem Darwish, Mohamed Al-Badrashiny, Mona Diab, Jan Niehues
The 20th International Conference on Spoken Language Translation Long Paper
TLDR:
We present the ACL 60/60 evaluation sets for multilingual translation of ACL 2022 technical presentations into 10 target languages. This dataset enables further research into multilingual speech translation under realistic recording conditions with unsegmented audio and domain-specific terminology,
You can open the
#paper-IWSLT_2
channel in a separate window.
Abstract:
We present the ACL 60/60 evaluation sets for multilingual translation of ACL 2022 technical presentations into 10 target languages. This dataset enables further research into multilingual speech translation under realistic recording conditions with unsegmented audio and domain-specific terminology, applying NLP tools to text and speech in the technical domain, and evaluating and improving model robustness to diverse speaker demographics.