Morlet小波
计算机科学
复调
小波
语音识别
小波变换
信号(编程语言)
常数Q变换
离散小波变换
人工智能
声学
物理
程序设计语言
作者
Neeraj Kumar,Raubin Kumar,Govind Murmu,Prabira Kumar Sethy
标识
DOI:10.1016/j.micpro.2020.103612
摘要
In this paper, the extraction of melody for polyphonic music is discussed. The method proposes a simple but effective modification in the Morlet wavelet as a modified Morlet wavelet (MMW). The pitch detection concept is applied to the transformed wavelet music signal for primary melody estimation and extraction. The proposed method is then compared with three other methods reported in the literature (the method focused on Short Time Fourier Transform, multi-resolution Fast Fourier Transform, and Morlet Wavelet). The percentage gross pitch error (GPE) parameter is considered for purposes of comparison. In comparison, it is observed that the proposed approach has the lowest GPE. The simulation results also present a lower percentage error in the predominant pitch frequency estimation for the proposed method.
科研通智能强力驱动
Strongly Powered by AbleSci AI