基于非线性动力学的乐器分类方法

本文基于非线性动力学理论,对不同乐器产生的音频时间序列进行了相空间重构,通过分析各类乐器的递归特性,提出了一个新的定量递归参数——密集度,它能够描述管乐器、弦乐器和键盘乐器在相空间中的差异,然后将密集度与传统的音色特征相结合,提出一种乐器分类方法,并将其应用于不同的分类模型.实验表明,本文所提的方法使三类乐器家族的分类准确率提高了4%～7%,单个乐器的分类准确率提高了3%左右.

The Musical Instrument Classification Algorithm Based on Nonlinear Dynamics

In this paper,the phase space reconstruction of audio time sequences producing by different instruments is discussed based on the nonlinear dynamic theory.The dense ratio,a novel quantitative recurrence parameter,is proposed to describe the difference of wind,string and keyboard instruments in the phase space by analyzing the recursive property of every instrument.In addition,a method of musical instrument family classification is developed using several classification models by combining the dense ratio with traditional timbre features.The experiments indicate that the accuracy of the proposed method is improved by 4%-7% and 3% in the instrument family classification and individual instrument classification,respectively.