Audio Classification Algorithm Based On GMM And SVM

With the development of computer, network and communication technology, multimedia data, which include image, video and audio, becomes primary information media in the field of information processing. Audio data is an important part of multimedia applications.Raw audio data is non-semantic and non-structured binary data stream which lacks semantic content description and structured organization. These characteristics of audio data bring much difficulty for audio information processing and analysis deeply. How to extract the structure information and semantic content from raw audio data is a key problem for audio structure parsing with audio assistance. Audio classification is key technology for solving such problems, and is the basis of audio structured. Incorporating the past research fruits, this thesis studies audio classification that solves the following problems, such as audio feature analysis and extraction for audio classification, GMM-based audio classifier and SVM-based audio classifier. At last, a system based on gaussian mixture model (GMM) statistical parameters and support vector machine (SVM) for audio classification is proposed and evaluated.The research work can be concluded as follows:(1) Audio features are described and evaluated.(2) GMM-based audio classifier and SVM-based audio classifier are studied, and the performances of these two classifiers are compared.(3) A voice activity detector algorithm using support vector machine is proposed.(4) A system based on gaussian mixture model (GMM) statistical parameters and support vector machine (SVM) for audio classification is proposed. The system combined the robustness of generative model and the powerful classification of discriminative model to improve the performance and robustness of classification.