I want to try creating a jQuery slideshow using simple voice commands like "next" or "previous". Is there a way to use Google's voice recognition? I know about Chrome's x-webkit-speech, but I have to click a button to use it.
I tried MIT's WAMI, but I found it slower and less accurate than Google's speech cognition.
As of today this now exists, for Chrome: http://chrome.blogspot.co.uk/2013/01/hello-browser.html
(api doc)
For an easy way to do this with JavaScript, check out annyang, which is a library that makes dealing with speech recognition super-easy.
The issue is what will capture your user's speech and send it some speech server for processing. x-webkit-speech is built into the browser and the browser uses the PC's sound card to capture speech. Javascript/Jquery in a browser alone cannot access the hardware devices. I do not know if Chrome exposes any x-webkit-speech objects in the DOM so you can do capture through Javascript.
That said, people have reverse engineered the Google speech api and used it from code. See http://mikepultz.com/2011/03/accessing-google-speech-api-chrome-11/ for an example.
I found this, I have not personally used it do I don't know if it is entirely what you are looking for but take a look.
http://captico.com/introducing-speech-2-text-api-by-google/2011/03