Combined Supervised and Unsupervised Approaches for Automatic Segmentation of Radiophonic Audio Streams

Author

Richard, Guilhem ; Ramona, Mathieu ; Essid, Slim

Author_Institution

GET-ENST, Paris, France

Volume

2

fYear

2007

fDate

15-20 April 2007

Abstract

Speech/music discrimination is one of the most studied topics in the domain of audio data segmentation. In this paper, we propose and evaluate a novel method that includes feature selection and a combined supervised and unsupervised strategy for audio streams segmentation. A number of alternatives solutions for each component are assessed and the optimized system is compared to the approaches proposed in the framework of the ESTER campaign.

Keywords

audio signal processing; feature extraction; ESTER campaign; audio data segmentation; automatic radiophonic audio stream segmentation; feature selection; speech-music discrimination; supervised approaches; unsupervised approaches; Cepstral analysis; Feature extraction; Hidden Markov models; Kernel; Mel frequency cepstral coefficient; Radio broadcasting; Speech; Streaming media; Support vector machine classification; Support vector machines; Speech/Music discrimination; audio segmentation; novelty detection;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech and Signal Processing, 2007. ICASSP 2007. IEEE International Conference on

Conference_Location

Honolulu, HI

ISSN

1520-6149

Print_ISBN

1-4244-0727-3

Type

conf

DOI

10.1109/ICASSP.2007.366272

Filename

4217445