YouTube Will Have Machine-Generated Automatic Captions

Google is planning to launch machine-generated automatic captions on its YouTube. Such opportunity will be available during three years. Google will use speech-recognition algorithms integrated in Google Voice in order captions for all videos could be generated automatically.

According to the Google’s blog 20 hours of video are downloaded to YouTube almost every minute. Auto-caps is aiming to get a big amount of videos quickly. Moreover captions is a good helping hand in getting access of video content as well as searching and giving a possibility to emphasize exact sections of videos.

But unfortunately only limited number of partner channels will provide auto-caps, they are U.C. Berkeley, Stanford, MIT, Yale, UCLA, Duke, UCTV, Columbia, PBS, National Geographic, Demand Media, UNSW and most Google and YouTube channels. According to Google the technology needs additional elaborations and also the company is wating for responses from viewers and video owners before a broader rollout.

At present Google installed automatic caption timing that will simplify the process of creating captions manually. Video creators make and download a text file with all the words in the video. Google has such feature as automatic speech recognition (ASR) technology that identifies the spoken words and creates the captions. Google believes that due to the technology it will be possible to load more captioned videos to the site as the cost is reduced.

Category: Technology News