Hi, my name is Adi. i'm currently working on a speaker recognition
program using delphi. i already got the way to read wave file. but i'm
losing the concept of speaker recognition. i want to ask about the
ideal apeaker recognition program. after i read the wave file, what
should i do next? frame blocking or windowing? then do the fft? or fft
first? and should i use lpc or preemphasize? what is the use of fft,
ifft, lpc, dwt ? i want to have coeficients that i can use at the
neural network so my program could recognize the speaker who spoke the
word. and how could i get the voiceprint?

sorry if i ask to much question. i'm a newbie in this area of research
and i had to get my paper done (the program too). i'm really
appreciate any help given. thank you very much.

regards,
adi Start a New Thread