Direct Speech Translation for Automatic Subtitling

arXiv (Cornell University)

Abstract

Automatic subtitling is the task of automatically translating the speech of audiovisual content into short pieces of timed text, i.e. subtitles and their corresponding timestamps. The generated subtitles need to conform to space and time requirements, while being synchronised with the speech and segmented in a way that facilitates comprehension. Given its considerable complexity, the task has so far been addressed through a pipeline of components that separately deal with transcribing, translating, and segmenting text into subtitles, as well as predicting timestamps. In this paper, we propose the first direct ST model for automatic subtitling that generates subtitles in the target language along with their timestamps with a single model. Our experiments on 7 language pairs show that our approach outperforms a cascade system in the same data condition, also being competitive with production tools on both in-domain and newly-released out-domain benchmarks covering new scenarios.

Authors 6

1. Sara Papi

   University of Trento · Fondazione Bruno Kessler

   Affiliation as printed

   Fondazione Bruno Kessler , Trento , Italy

   University of Trento , Trento , Italy

2. Marco Gaido

   University of Trento · Fondazione Bruno Kessler

   Affiliation as printed

   Fondazione Bruno Kessler , Trento , Italy

   University of Trento , Trento , Italy

3. Alina Karakanta Aachen

   Leiden University

   Affiliation as printed

   Leiden University Centre for Linguistics , Leiden , The Netherlands

4. Mauro Cettolo

   Fondazione Bruno Kessler

   Affiliation as printed

   Fondazione Bruno Kessler , Trento , Italy

5. Matteo Negri

   Fondazione Bruno Kessler

   Affiliation as printed

   Fondazione Bruno Kessler , Trento , Italy

6. Marco Turchi
   Affiliation as printed

   Zoom Video Communications , Karlsruhe , Germany