SPEECH - MUSIC DISCRIMINATION DEMO
VERSION 1.0

DOWNLOAD

You can download the executable version of the demo here.
IMPORTANT: If you do not have Matlab v. 7.0.0 installed on your Pc, you will have to download the Matlab Component Runtime installer here .
Visit the Documentation-->Installation section for more information on installation issues.

We also provide a m-file for manual segmentation. The file can be downloaded here. When you run the m-file in the Matlab workspace, you will be asked to give:

  1. The number of segments.
  2. For each segment: (a) The end of the segment (in seconds) and (b) The label of the segment ('S' for speech and 'M' for music).
The argument of the function is the filename of the audio file (wav file) in which the audio stream is stored. When the manual segmentation process is completed, a mat file is created, named _true.mat.
You can use the produced mat file to compare the demo's results with the manual segmentation results and therefore to produce a confusion matrix. To do this, simply place the produced m-file in the same directory with the audio (wav) file.

Documentation for the demo is available for both the installation and the execution of the demo. For more information about the proposed algorithm please refer to the papers that are listed in the "References".page.

For more questions or comments see the contact information .