Song-level features and SVMs for music classification

Searching and organizing growing digital music collections requires automatic classification of music. Our system for artist and genre identification uses support vector machines to classify songs based on features calculated over their entire lengths. Since support vector machines are exemplar-based classifiers, training on and classifying entire songs instead of short-time features makes intuitive sense. We model songs as single Gaussians of MFCCs and use a KL divergence-based kernel to measure the distance between songs. This system placed first in both the audio genre and artist identification competitions at MIREX with classification accuracies of 72.45% and 78.81%, respectively.