ROSETTA’s final demonstrator

This demonstrator shows, in an extract of TV news headlines, the results of the research carried out in the technical modules, at the end of the ROSETTA project (November 2021): 

  • French subtitles automatically generated by LISN (François Buet and François Yvon)
  • multilingual subtitles automatically generated by SYSTRAN (Elise Bertin-Lemée and Josep Crego)
  • avatar signing in LSF : 
    • corpora recorded by MOCAPLAB (Louis Clavel and Boris Dauriac)
    • translation algorithm from French to LSF via AZee by SYSTRAN (Elise Bertin-Lemée) and LISN (Michael Filhol)
    • AZee annotation, coordination and LSF revisions by LISN (Annelies Braffort, Camille Challant, Claire Danet, Michael Filhol)
    • 3D avatar generation by MOCAPLAB (Boris Dauriac)

The following steps also contributed to this demonstrator: 

  • video production: MOCAPLAB (Boris Dauriac)
  • sound transcription: LISN (François Buet)
  • FTV player integration: Holken Consultants (Youssra Gouram)