- From: Silvia Pfeiffer <silviapfeiffer1@gmail.com>
- Date: Sat, 21 Nov 2009 07:44:42 +1100
- To: Geoff Freed <geoff_freed@wgbh.org>
- Cc: HTML Accessibility Task Force <public-html-a11y@w3.org>
I couldn't agree more. I think it's a first step and will need continuous improvement - in particular since automated speech recognition technology in general is not perfect (and may never be). But I expect that it's just a first step of Google - probably a typical "beta" - and they will hopefully continue to improve it. Feedback is certainly encouragement for that. I'm sure Ken Harrenstien who attended the Video Accessibility Workshop ahead of TPAC would agree. Cheers, Silvia. On Sat, Nov 21, 2009 at 2:18 AM, Geoff Freed <geoff_freed@wgbh.org> wrote: > > I agree that this is exciting news for captioning, and I applaud Google for > once again leading the way. Having a company as influential as Google pay > attention and devote significant resources to accessibility is a very > important breakthrough. > > However, I also urge everyone interested in captions to give this new > approach and software a thorough test. While auto-caps is not yet available > for public consumption, auto-timing is. I’ve been testing it today and have > found that it’s reasonably accurate in some situations but has a lot of room > for improvement. When auto-caps becomes available for everyone to use, I > urge everyone here to give it a good workout and provide useful feedback to > Google. As Google admits, auto-caps’ output is not now perfect. But it is > crucial that its output *becomes* perfect in order for it to be optimally > useful for both viewers and authors. > > Geoff/NCAM > > > > On 11/19/09 10:44 PM, "Silvia Pfeiffer" <silviapfeiffer1@gmail.com> wrote: > > Hi all, > > Thought you should know: Google/YouTube continues to paint the way - > of follow John Foliot here. :) > > http://googleblog.blogspot.com/2009/11/automatic-captions-in-youtube.html > > They provide automated captioning using speech recognition, but only > for specific publishers right now. > > What's more interesting is that they allow for the provisioning of > transcripts and automatically create the time-alignment. This is very > similar to a service that John Foliot explained in the Video > Accessibility Workshop before TPAC. > > I'm very excited about this - now they just need to start doing > textual audio descriptions and hook them up to screenreaders and we > have made a huge step forward in the online video world. > > Now we just need to do the same for HTML5. :-) > > Cheers, > Silvia. > > >
Received on Friday, 20 November 2009 20:45:42 UTC