Abstract: In music retrieval methods based on audio fingerprints,a large database is required in order to compare with the fingerprints leading to low retrieval efficiency.In this paper,we proposed a method for index compression using a compressed suffix array,in order to overcome the disadvantage of large memory cost with full-text indexing.The proposed method compresses data sequences lossless by run length encoding,mainly taking advantage of the fact that the repetitivecharacters occur frequently in higher bits of the sorted audio fingerprint data.The experimental results show that the proposed method,compared with the conventional method,only needs 30% of the space of an audio fingerprints database for a music database consisting of 12000 songs,and around 80% of the index space for a database of 2000 songs.Moreover,theentire space cost is reduced to around 60%,compared with the method based on the suffix array.

Key words: Audio fingerprint,Compressed suffix array,Index compression,Run length encoding,Vertical code