Towards a Fast and Efficient Match Algorithm for Content-Based Music Retrieval on Acoustic Data

In this paper we present a fast and efficient match algorithm, which consists of two key techniques: Spectral Correlation Based Feature Merge(SCBFM) and Two-Step Retrieval(TSR). SCBFM can remove the redundant information. In consequence, the resulting feature sequence has a smaller size, requiring less storage and computation. In addition, most of the tempo variation is removed; thus a much simpler sequence match method can be adopted. Also, TSR relies on the characteristics of Mel-Frequency Cepstral Coefficient(MFCC), where the precise match in the second step depends on the first step to filter out most of the dissimilar references with only the low order MFCC feature. As a result, the whole retrieval speed can be further improved. The experimental evaluation verifies that SCBFM-TSR yields more meaningful results in comparatively short time. The experiment results are analyzed with a theoretical approach that seeks to find the relation between Spectral Correlation(SC) threshold and storage, computation.