7th International Conference on Spoken Language Processing September 16-20, 2002 Denver, Colorado, USA

A Sound Source Classification System Based on Subband Processing

Oytun Turk, Omer Sayli, Helin Dutagaci, Levent M. Arslan

Bogazici University, Turkey

A system for classification of audio signals containing speech, music, noise and silence is proposed. Appropriate subband processing is applied for the characterization of each sound source. The algorithm operates in four steps to classify the contents of a given audio signal. The acoustical parameters and statistical measures to be used in the classification process are obtained via an off-line training procedure. The starting and finishing instants of the acoustical events are labelled in the silence/onset detection stages. Acoustical parameters of the given signal are extracted, analysis of variance and classification using the LBG algorithm is performed by generating codebooks of acoustical vectors. Experimental work is carried out on a database containing speech, music, noise and silence. The experiments demonstrate that the system achieves 88% classification success on the average when the sound sources are non-simultaneous.

Full Paper

Bibliographic reference.  Turk, Oytun / Sayli, Omer / Dutagaci, Helin / Arslan, Levent M. (2002): "A sound source classification system based on subband processing", In ICSLP-2002, 641-644.