Main Melody Extraction Using the Auditory Scene Analysis for the Humming Music Retrieval

The query-by-humming system based on waveform files which are usually a mixture of strong accompaniment and singing voice will be the main trend of the future. The accompaniment and the singing voice have many overlaps in spectrum, which leads to a big interference to singing voice signal. So how to accurately extract the main melody of the singing voice is a very valuable research topic in the case of containing accompaniment for waveform files. In this paper, main melody extraction algorithm using the auditory scene analysis for the humming music retrieval with the improved dynamic time warping algorithm is proposed. Experiment results show that the retrieval correct rate is improved by 6.2% compared with the existing traditional humming music retrieval system.