We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:


References & Citations

DBLP - CS Bibliography


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computation and Language

Title: Joint Generation of Captions and Subtitles with Dual Decoding

Abstract: As the amount of audio-visual content increases, the need to develop automatic captioning and subtitling solutions to match the expectations of a growing international audience appears as the only viable way to boost throughput and lower the related post-production costs. Automatic captioning and subtitling often need to be tightly intertwined to achieve an appropriate level of consistency and synchronization with each other and with the video signal. In this work, we assess a dual decoding scheme to achieve a strong coupling between these two tasks and show how adequacy and consistency are increased, with virtually no additional cost in terms of model size and training complexity.
Comments: Accepted at IWSLT 2022
Subjects: Computation and Language (cs.CL)
Cite as: arXiv:2205.06522 [cs.CL]
  (or arXiv:2205.06522v1 [cs.CL] for this version)

Submission history

From: Jitao Xu [view email]
[v1] Fri, 13 May 2022 09:12:42 GMT (71kb,D)

Link back to: arXiv, form interface, contact.