- Image via Wikipedia
In my end-of-day RSS reader catch-up, I stumbled across an article outlining some new YouTube services. Essentially, YouTube/Google is now applying automatic speech recognition technology they’ve been perfecting withing Google Voice, to close-caption YouTube videos. They’re also a providing mechanisms for “auto-timing” self-provided captions uploaded as simple text files. These are exciting new features that will greatly enhance the ability to find appropriate video content as well as become more accessible in general. I hope to see some of these technologies made reasonably available for implementation outside of YouTube. I know in the school systems, YouTube is often entirely blocked. It would be fantastic to see sites such as TeacherTube able to leverage the technology to increase their reach. Be sure to read the full post at Google as there are a few other tidbits such as caption translation that are demonstrated.
2 Comments
James: You should consider sharing a presentation about these functions at OTA in February. I bet there would be a lot of interest. I’m eager to see if we can utilize any of this for k12online this year. Since we’re not a registered 501.c3 we don’t qualify for YouTube’s nonprofit program, and I think you have to be in that or a registered “partner” to publish videos longer than 10 minutes. Any suggestions for us? We’re using DotSub but it doesn’t have this automated option. The download and upload option for caption files is great too!
I’d definitely consider OTA. Unfortunately, I believe it coincides with an annual meeting I must attend.
As for the nonprofit program, perhaps if k12online partnered with a recently incorporated 501.c3 that had some synergistic goals. Or the group could attempt to get added to the Partner Program through the “reputation” method. Post a series of acceptable 10 minute or less videos from past events and get recognition of the solid management of the account. It’s unfortunate there isn’t a simpler process to meet some of these educational goals.
I think the options to download and upload for the auto-timing were my favorite. It’s a topic I’ve heard on a number of occasions in past OTA and MoodleMoot conferences. I know first hand that the ASR for Google Voice has limitations. I can only guess what it might be like with average uploaded video content.