Voice Recognition

[asaf123456]

Hi All

I want to record several words once in my voice using the cellular's microphone (example: "house", "in", "garden", "the").

after recording each word individualy into a file / data base, i want to say a sentence containing those words - for example i want to say: "garden in the house".

I want to record the sentence and break it into words (i will know where each word ends according to the quiet between each 2 words) and than i want to adjust the word i seperated to those i recorded earlier.

when a match will be found, the text of the word will be presented on the screen.

does anyone knows how to address this problem ??

thanx

Asaf

[wizard_hu_]

Generally voice/speech recognition is a non-trivial task, and although CSpeechRecognitionUtility might be existing,

This item is not part of the S60 SDK

thus you will likely have to develop your own method. Recording itself is easy CMdaAudioInputStream can provide you -say- a PCM16 stream, there might be examples in the Wiki about it and/or on FN main site (check www.forum.nokia.com Documentation/Symbian C++ on the left, Multimedia).
Then try searching for speech recognition or voice recognition (with Google for example), since the algorithms you will need are not OS-specific. Most methods start with changing to the frequency domain, then do some statistical analysis.