Speaker diarization and speech recognition in the semi-automatization of audio description: An exploratory study on future possibilities?

Abstract

This article presents an overview of the technological components usedin the process of audio description, and suggests a new scenario inwhich speech recognition, machine translation, and text-to-speech, withthe corresponding human revision, could be used to increase audiodescription provision. The article focuses on a process in which bothspeaker diarization and speech recognition are used in order to obtain asemi-automatic transcription of the audio description track. Thetechnical process is presented and experimental results are summarized.