max/msp/jitter Sound - Speech segmentation/recognition/analysis

Question

I am trying to develop a speech / sound recognition program which extracts some useful data such as sound is concerned. For instance .. fundamental frequency / MFCC / Centroid etc. Speech is usually segmented in frames of 20 to 30 ms, and the window analysis is shifted by 10 ms.

I would like to find a patch / object or some useful advice on how can I achieve a window segmentation with the frames, the shift, the step that I prefer for a sound analysis-segmentation.

Does anybody know a way to do this ?

Answer 1

You could try

• ejs: http://www.e--j.com/index.php/what-is-zsa-descriptors/
• Gabor objects (gbr.*) in the FTM library : http://ismm.ircam.fr/ftmco/