Low-rank audio signal classification under soft margin and trace norm constraints

We propose an algorithm to do speech/non-speech classification based on low-rank matrix representative audio data. Conven-tionally, the low-rank matrix data can be represented by a vec-tor in high dimensional space. Some learning algorithms are then applied in such a vector space for matrix data classifica-tion. Particularly, maximum margin classifiers, such as support vector machine (SVM) etc. have received a lot of attentions due to their effectiveness. In this paper, we classify the data directly in the matrix space. Our methodology is built on recent stud-ies about matrix classification with the trace norm constrained weight matrix and SVM's large-margin linear discrimination principle. The resulting low-rank SVM is then designed to max-imize the margin between classes whilst minimizing the com-plexity of the classifier in both original and low-rank space. We compared our proposed algorithm with SVM and other state-of-the-art matrix classification methods. Experimental studies on real life audio signal classification show the effectiveness of our algorithm.