Segmentation

Trained (i.e. we know how many and what speakers)

or untrained (unknown number and voice of speakers)

Separate takes, different content, parts of programmes

e.g. listen only to segment starts instead of having to scan through a whole track

Tzanetakis, G., & Cook, P. (1999). Multifeature audio segmentation for browsing and annotation. IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, 1–4. Retrieved from http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=810860

Process

The measurement of novelty and change from one texture window to the next can be a good indication of a change in section or content

But another useful technique is to look at the self-similarity matrix and extract the structure from it.

Foote, J. (2000). Automatic audio segmentation using a measure of audio novelty. Multimedia and Expo, 2000. ICME 2000. 2000 IEEE …, 1, 452–455. Retrieved from http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=869637