hi i am analysing speech signals. So i need a good transform which will not depend on the speaker.So, which transform is good? I heard that Fourier Mellin's transform is very good for this application.
Is it right?
If so if any of u have the m file of this transform please upload it.
Many features are used for speech analysis. For speech recognition, for example, you can use MFCC (Mel Frequencies Cepstral Coefficients). For speech analysis, significants features are: Cepstral Coefficients, Linear prediction coefficients, PLP and Rasta-PLP coefficients. Some research are focused on the use of combined parameters.