Friday, February 17, 2017

One Billion Videos Captioned.. sort of

While YouTube is bragging about have one billion captioned videos, critics are pointing out that many of the automatically generated captions are wildly inaccurate. It's been eight years since Google added the automatic speech recognition designed to generate captions and the company says it's accuracy is up by 50 percent. But in the anouncement about the number of captioned videos, Google Product Manager Liat Kaver admits the program is not where the company wants it to be:
A major goal for the team has been improving the accuracy of automatic captions — something that is not easy to do for a platform of YouTube’s size and diversity of content. Key to the success of this endeavor was improving our speech recognition, machine learning algorithms, and expanding our training data. There were limitations with the technology that underscored the need to improve the captions themselves. Results were sometimes less than perfect, prompting some creators to have a little fun at our expense!
Read the full announcement from YouTube here.