Speech processing has always had a major role to play in making interactive media more accessible. Now, as Apple, Google, Netflix, Sony and others square off to define how the Internet, search, streaming and social media all coalesce on the tube, Nuance Communications and the National Center for Accessible Media are stepping up to build a system that assesses the ability to render spoken words as accurate captions.

Nuance will be leveraging the speech recognition and transcription capabilities of Dragon NaturallySpeaking. It plans to develop and deliver customized language processing, data analysis, and benchmarking tools for the Project. The objective is to build a “prototype tool that will enable tracking of audio and captioning, comparing the spoken word with the caption output and rating the accuracy levels based on error type and severity.”

Accurate transcription of video soundtracks has importance that transcends living rooms and TVs. It is foundational to a variety of initiatives that support search, analytics and real-time translation to support defense initiatives, corporate videoconferencing and video calling (like Apple FaceTime).

The Public Broadcasting Services’ Boston affiliate, WGBH-TV, established the National Caption Center back in 1972. and has been home to The Carl and Ruth Shapiro Family National Center for Accessible Media (NCAM) for a number of years. More info on the Media Access group can be found at access.wgbh.org.

Categories: Articles

Tags: Video, Dictation, Analytics, speech processing