<br><br><div class="gmail_quote">On Thu, Mar 27, 2008 at 4:21 AM, Shawn Rutledge <<a href="mailto:shawn.t.rutledge@gmail.com">shawn.t.rutledge@gmail.com</a>> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div class="Ih2E3d">On Wed, Mar 26, 2008 at 3:18 PM, saurabh gupta<br>
<<a href="mailto:saurabhgupta1403@gmail.com">saurabhgupta1403@gmail.com</a>> wrote:<br>
> 1)Speech Recognition Method:<br>
><br>
> i have already developed a speech recognition method on 8-bit<br>
> microcontroller. i used the digital filters(chebyshev II) and correlation<br>
> method to match the spoken word with that already stored in memory. i am<br>
> further planning to improve it using Hidden markov model(HMM) through Linear<br>
> Predictive Coding (LPC) and fast fourier transform (FFT). A success rate of<br>
> 90% can be achieved using this method.<br>
<br>
</div>I'm impressed you managed to do that on an 8-bit microcontroller!<br>
<br>
Free software for speech recognition has been slow to develop in the<br>
past. Now there are a couple:<br>
<br>
<a href="http://freshmeat.net/search/?q=speech+recognition&section=projects&Go.x=0&Go.y=0" target="_blank">http://freshmeat.net/search/?q=speech+recognition&section=projects&Go.x=0&Go.y=0</a><br>
<br>
E.g. there is Sphinx and Julius, and there is also an HMM toolkit<br>
which might be useful (or not).<br>
<br>
Do you expect to achieve better performance than those?<br>
</blockquote></div><br>hello sir, <br>right now i cannot say surely that whether i will be able to achieve a better result than those already developed like Sphinx or HMM toolkit, but i can try to embed the method of Sphinx and julius into open moko. as a student i will try to understand the concept used in the methods developed previously and try to modify it accordingly so as to get it fixed in open moko.<br clear="all">
<br>-- <br>Saurabh Gupta<br>3rd year,<br>Electronics and Communication Engg.<br>NSIT,New Delhi