Music/voice separation based on the multi-repeating structure of Mel-frequency cepstrum coefficients

For the poor adaptability of the original repeating pattern,an improved music separation method of multirepeating structure of Mel-Frequency Cepstrum Coefficient(MFCC) was proposed.Firstly,the MFCC coefficient matrix(39-dimensional data) of the music signal was extracted;then the cosine characteristic was applied to the count of similarity matrix of MFCC,and putted the fragments with consistent similarity together,next built different repeating patterns for groups with different,thereby the spectrums of the background music and vocal were separated combined with ideal binary masking(IBM),the corresponding time domain signals were obtained by inverse Fourier transform;finally,the improved method was tested on the music database of different types and length,and the separation results were compared with repeating method of Rafii and the non-negative matrix factorization based on flexible framework method of Ozerov.The experimental results showed that the separation performance of improved method was improved about 3 dB,the performance of music with melody changed larger was significantly improved,thus verifying that that the improved method was an effective music separation algorithm and more