Audio Classification and Segmentation Based on Support Vector Machines

Audio classification and segmentation are an important access to extract audio structure and content, and are a basis for further audio/video retrieval and analysis. Support vector machines (SVM) is a valid statistic learning method. In this paper, the work on audio classification based on SVM is presented. Five audio classes are considered in this paper: silence, noise, music, pure speech and speech over background sound. Three smooth rules are present- ed and applied in the final segmentation. The performance of SVM on audio classification is evaluated. The effective- ness of some new proposed features is also evaluated. Experiment results show that SVM performs very well for au- dio classification and segmentation accuracy is good with the proposed three smooth rules.