How to Demo Chrome 11′s Speech Recognition Feature

Posted on 23. Mar, 2011 by in Tips

flattr this!

With Chrome 11 hitting the Beta Channel, a number of new improvements to the V8 JavaScript engine and GPU-accelerated 3D CSS have been added. But what sticks out the most in this release is a new API that allows for speech input. While there aren’t many places on the web you can use this right now, there is a demo at the Google-hosted HTML5 Rocks site here.

You need to have Chrome 11 in order for this to work. Once you get to the HTML5 Rocks site, you’ll notice a field with a little microphone. Just click the mic icon and speak whatever it is you wish. In a few moments you’ll see that the word(s) appear in the text box.

chromespeechinput

This technology is similar to the Google Voice Search that the company has for smartphones. But instead of being trapped in a mobile app on a phone, the speech input API could be used for a number of things that creative developers will be able to come up with.

What uses could you see the speech input API being used for on the web?

via Google Chrome Blog, DownloadSquad

  • services sprite
  • services sprite
  • services sprite
  • services sprite
  • services sprite
  • services sprite
  • services sprite
  • services sprite
  • services sprite

Related posts:

  1. Using the Speech Attribute Feature in Chrome 7
  2. Why Use Chrome’s Speech Input API When You Can Speechify?
  3. Dev Channel of Chrome to Get Feature That Protects Against Malicious Files

Tags: , , , , , ,

4 Responses to “How to Demo Chrome 11′s Speech Recognition Feature”

  1. Matthew

    23. Mar, 2011

    Does the speech recognition computation happen in the browser code, or do they send the voice bits to their Google servers which translates it into text and then send it back to the browser?

    The latter is how voice recognition works on Android, and why it only works when you have an active internet connection.

    It seems a little weird to me to build an html5 standard that requires server side computation. How is a non-profit open source browser going to fund the massive server load needed, not to mention the R&D needed to develop a voice-to-text translator? Google’s stuff is proprietary.

  2. Daniel Cawrey

    23. Mar, 2011

    It needs the work of a server-side translation tool as described in this blog post. True, that means the API would not be a open standard, but many Google APIs are not anyways and this is especially true when a platform uses their own resources.

  3. bob

    23. Mar, 2011

    peach reckon mission is working grate sew far!

  4. morocarlo

    23. Mar, 2011

    If you wants speech recognition in your website yust put:

    //not neccesary

    and it’s works (only with chrome 11) :P

    this is my italian post:
    http://www.chromeos.eu/2011/03/23/chrome-11-ci-capisce-ecco-come-fargli-capire-litaliano/

    p.s. sorry for my english

Leave a Reply

Name

Email

Website

Comment