Can the same words at different time from a single person be the same or approximatly the same, using sound processing on mathlab?

if so please would u give me the algorithm using mathlab?

[Walter Roberson]

Are you asking how to do speech recognition?

[Walter Roberson]

Or maybe you are trying to do speaker recognition (voice recognition) ?

[solomon adenew]

ya, in voice recognition

Answer 1

[Youssef Khmou]

adenew,

There is discipline in signal processing and audio engineering specialized with this type of application, many methods are applied such as Hidden Markov Models, Genetic algorithm, Fractional Fourier transforms, and so on, The easy way is to use Discrete Fourier transform (fft) or Correlation between the two signals , total similarity gives you correlation coefficients of [1 1;1 1]

[solomon adenew]

really, thanks

[Walter Roberson]

Note: individuals pronounce the same word different ways at different times; and the pronunciation of any given basic sound is fairly dependent upon context of the surrounding sounds.

Most of the United States pronounce the three words "Mary", "merry", and "marry" the same way, but North-East United States, especially near Pennsylvania, pronounce the three differently. The part of Canada that I come from is north of that portion of United States, and I too pronounce them distinctly (and can hear the difference in pronunciation), but the difference in sound is quite difficult to describe. All three of them are officially considered to use the same sounds, but the speakers in the area modify that sound according to the other sounds close by.