Quick audio retrieval using multiple feature vectors

This paper concentrated on the content-based audio retrieval and proposed a quick audio search algorithm. In the previous work, the time-series audio search method with the upper bound proof was generally used. However, the searching speed of a time-series search method is very poor at real time. Therefore, in present paper the preprocessing stage is introduced for this defect. And to enhance the search accuracy we use the multiple feature vectors of audio signal. Through the experiment, the results of retrieval performance are presented.