基于节奏的鼓声识别

从鼓声能量与鼓声循环两个角度出发,提出了一种从多音信号中识别鼓声的方法。充分利用鼓声能量高于一般谐波乐器的特性,通过对重音信号、NAE拍谱、重音拍谱有效组合的方式,将鼓声的起点在识别函数中突出反映出来,从而根据节奏信息来识别鼓声的位置。实验结果表明,算法在鼓声识别的正确率和召回率上,较常用的K.Yoshii时频域模板迭代法和A.Zils时域模板迭代法有大幅度提高。

Drum Sounds Recognition Based on Rhythm

This paper describes an approach to detect drum sounds from polyphonic music based on drum energy and drum loop. Considering the energy of drum is much higher than that of harmonic instruments, the possible drum onsets were indicated by efficiently combining accent signal, NAE beat spectrum, and accent beat spectrum into a clearer recognition curve. The drum segment was then recognized using the rhythm information. The experiments show that the proposed al- gorithm improves the accuracy and the recall rate compared to the common methods using template iteration in either timefrequency domain or time domain.