Cover song detection: From high scores to general classification

Suman Ravuri, Daniel P.W. Ellis
2010 2010 IEEE International Conference on Acoustics, Speech and Signal Processing  
Existing cover song detection systems require prior knowledge of the number of cover songs in a test set in order to identify cover(s) to a reference song. We describe a system that does not require such prior knowledge. The input to the system is a reference track and test track, and the output from the system is a binary classification of whether the reference/test pair is either from a reference/cover or reference/non-cover. The system differs from state-of-the-art detectors by calculating
more » ... ltiple input features, performing a novel type of test song normalization in order to combat against "impostor" tracks, and performing classification using either a SVM or multi-layer perceptron (MLP). On the covers80 test set, the system achieves an equal error rate of 10%, compared to 21.3% achieved by the 2007 LabROSA cover song detection system. Index Terms-Cover songs, music information retrieval.
doi:10.1109/icassp.2010.5496214 dblp:conf/icassp/RavuriE10 fatcat:tyfmzodxmnfrripf5hapch4h64