Feature extraction for speech and music discrimination | IEEE Conference Publication | IEEE Xplore