a voice tag is entered via the microphone to the PC.In the PC matlab regonizes the speech and compares with previously saved samples.after the comparison the compared voice tag is converted to text {in matlab enviornment} and text is transmitted to a wireless transciever module for further stages.
{the speech recognition is the first stage}.
although i hve written the code for voice recognition in matlab.
the problem is i m not able to convert voice to text in matlab.
Voice signals will have discrete values. For example take a sine wave, sample it and quantize it. What you will have is of the form of no.s representing the signal amplitude. So basically numbers cant be converted into text . But you can do one thing, that is, after matching the sample with previously stored sample you can send pre-stored test to your next stage. The idea is to store a text string associated with every voice sample. I hope this helps.
hello dear,
m also doing speech recognition project..
m using spectrogram and dynamic time warping for speech recognition.
what technique are u using.
we can have some exchange of useful info regarding our projects.
Reagards
Atiq Baloch