Video: Microsoft Vista Speech Recognition Tested – Perl Scripting

February 14, 2007 on 5:47 pm | In funny, microsoft, recognition, software, speech, video, vista, windows | 4 Comments

So how good is the speech recognition software in the brand new Microsoft Vista anyway? Lets say for a second that you were able to successfully install MS Vista in your computer and you got it running. Now you want to test out the new features, one of them being this great speech recognition. Well, writting a simple document is just boring – so why not try to write an extremely short Perl script instead?

This Perl script would take around 20 seconds to write by typing it out.. So, seriously – It shouldn’t take more than 2, 3 or maybe even 5 minutes to write this Perl script using the Vista speech recognition, even while you are learning how to it, right?

Here is the video:

“delete if someone if someone delete the symbolism of the month of” – LOL

(Via: JWZ)

4 Comments »

RSS feed for comments on this post. TrackBack URI

  1. The person dictating is STUPID.

    Comment by VistaHater — February 14, 2007 #

  2. Fair enough as a piece of comedy but as a means of dissing vista speech recognition ? if you actually wanted to code why not use your hands…?or were you too busy rumbbing them together with glee at how smart your gonna look

    Comment by neveryoumind — February 15, 2007 #

  3. Really, the voice recognition isn’t that bad. He always says ’slektal’ and the computer usually knows that he means ’select all,’ which is pretty impressive.

    Comment by gvb — February 19, 2007 #

  4. This kind-of reminds me of the early Microsoft pen services for windows. Most people would get so frustrated trying to train the beast that they would opt for the little on-screen keyboard. I had an Itronix with a touch-screen and soon realized that the moer frustrated you got, the more difficult it was to write in a way that the limited recognition software could handle.

    The speech recognition appears to have the same problem. My guess is that the speech recognition simply filters the recorded speech through a fft algorithm and then tries to identify a sequence of phoneme patterns in the result that it uses as an index into a dictionary.

    Most of the serious speech recognition systems require training, a short procedure involving reading a list of words as they are displayed on the screen.

    The real problem here is that speech recognition has been under development for over 30 years and there have been products that worked quite well. If you had seen any of those applications in use, you wouldn’t call anything about this “impressive”.

    Comment by Niklaus — February 21, 2007 #

Leave a comment

XHTML: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Powered by WordPress with design by Borja Fernandez.
Entries and comments feeds. Valid XHTML and CSS. ^Top^