You too can YouTube

Content-sharing site YouTube has enabled people to share magic moments, material from entertainment archives and more informative video. But it’s a part of the internet that’s remained closed to some people. Until now. Edmund West reports

A major complaint among groups of disabled people, particularly those with sensory impairments, is that all too often the technological revolution leaves them behind.

One area where this has been evident is in content sharing: how, for instance, could a deaf person use­fully access the wealth of video content on YouTube?

After years of lobbying, auto-captions, also known as subtitles, are becoming a universal video feature. YouTube, having introduced them gradually, began extending them to its whole site on 4 March.

Its UK communications manager, Oliver Rickman, explained: “This is part of an ongoing process. One of the first steps we took was the development of a caption feature in 2008. In November of last year we released auto-captioning for a small, select group of partners.”

These first auto-captions were for a few educational videos; now they’re being released for all. That means that anyone uploading content to the site can request captions for it, which YouTube will provide.

The captions use automatic speech recognition (ASR) software to caption submitted videos on request. It works in the same way as Google Voice and Google’s speech-to-text voice search on mobiles. If a video has a CC button, click it and you’ll have a “transcribe audio” button for English captions and a “translate captions” button for other languages. You can even alter the font by clicking the “other settings” button.

Asked about the project’s next stage, Rickman said: “We’re working on improving our underlying ASR technology, with a focus on improving performance as well as expanding to other languages.

Our research includes a focus on expanding our ability to handle a wide variety of material (e.g. different speaking conditions and accents). The ultimate goal, though we don’t know how long it will take to get there, is to make this feature available for all videos in all languages.”

YouTube has a queue of people asking for their videos to be captioned early. Captioning content takes anything from an hour to a day.

With more deaf than Italian-speaking internet users, the impact of this technology could be vast. There were many calls from deaf people for this feature, Rickman said.

“Information access is core to our mission. We want to make information universally accessible and useful, and that certainly includes people with disabilities, such as blindness or low vision, colour deficiency, deafness, hearing loss and limited dexterity.”

Another group who could use captions are hearing people learning sign language.

Oliver admitted the system had its limitations. “Different accents, background noise, and audio quality levels present interesting challenges that we’re working to address as we improve our ASR technology.”

Maybe in the future, those improvements will extend to areas such as audio description for visually-impaired users.