Speech recognition: the solution for video subtitles
Video subtitling is a technique related to audiovisual content, especially in the film industry, consisting in the text display at the bottom of the image, during broadcasting of a program, like a movie. This technique, initiated by cinema – especially through silent movies to make characters talk – has been brought to television, where any types of programs like TV shows, documentaries or TV news can use it.
Live video subtitles – or teletext – consist of subtitling a program, in real time with a few seconds delay. All programs broadcast in live or not are concerned (TV news, football games, political debates, TV reality shows etc.) and addressed especially deaf and hard of hearing people. There is no reason that they do not have access to information.
How does it work?
All these tasks require a real teamwork. You need a “parrot” and a corrector.
The parrot, equipped with headphones and a microphone, repeats everything he has heard (speech of speakers on the screen, voiceover, etc. ).
Then everything is transcribed using a speech recognition software.
Finally, the text appears on the screen of the corrector who corrects mistakes (spelling, grammar, unknown words, etc.) before validating the text that appears on the screen of viewers.
This technique for video subtitles allows accessibility of audiovisual content for hard of hearing people.
- Video subtitling enables translation of an original version and to go global
More and more people prefer to watch their TV shows or movies in original version in order to not affect the video source and better enjoy game of actors. The dubbing is not always good and can alter perception we have of actors. For instance, Hugh Laurie, who has a deep voice, plays Dr. House with more drama than his French dubbing voice, good but more acute. Even if we don’t know the original language, we catch all the emotions of the actor and video subtitles are transmitting us his speech.
The Authôt transcription technology optimizes subtitle processing time
This speech recognition software is multi speakers, it allows to transcribe automatically audio from video into text with timecodes.
Thus this text file can be included directly into the video and subtitles appear. Then, the video becomes accessible. Video subtitles are in favor of digital accessibility!
Authôt: You speak, we write.