A Novel Music Retrieval System with Relevance Feedback

Although various researches have been conducted in the area of content-based music retrieval, however, few works have been done using relevance feedback for improving the retrieval performance. In this paper we introduce a novel content-based music retrieval system with relevance feedback. It enables users to search favorite music files by introducing the user as a part of the retrieval loop. In our system, we used a radial basis function (RBF) based learning algorithm and a method exploited both positive and negative examples to reweight feature components. Experiments evaluate the performance of the proposed approach and prove the effectiveness of our system.