Audio scene segmentation using multiple features, models and time scales.

Hari Sundaram, Shih-Fu Chang

Browse the full ICASSP paper archive.