Wavelet ridges for musical instrument classification

Abstract

The time-varying frequency structure of musical signals have been analyzed using wavelets by either extracting the instantaneous frequency of signals or building features from the energies of sub-band coefficients. We propose to benefit from a combination of these two approaches and use the time-frequency domain energy localization curves, called as wavelet ridges, in order to build features for classification of musical instrument sounds. We evaluated the representative capability of our feature in different musical instrument classification problems using support vector machine classifiers. The comparison with the features based on parameterizing the wavelet sub-band energies confirmed the effectiveness of the proposed feature.

Correlation and clusterisation of traditional Malay musical instrument sound using the I-KAZTM statistical signal analysis

“…Previously, there were some researchers that mainly focused on the polyphonic music signal for instrument classification [9][10][11][12][13][14][15].Past researchers like Deng et al [16] discussed feature analysis including the MPEG-7, statistical values of the mel frequency cepstral coefficient (MFCC), zero crossing rate (ZCR), root mean square (RMS), spectral centroid and flux using various machine learning techniques. In addition, Özbek et al [17] worked on the time-frequency energy of wavelet ridges using the Support Vector Machine (SVM) for automatic musical instrument classification. A further study was then conducted by Souza et al [18] by classifying drum sounds with the SVM algorithm using Line Spectral Frequencies (LSF).…”

Assessment of Laying Hens’ Thermal Comfort Using Sound Technology

“…Although they were initially designed to solve two-class classification problems, multi-class classifications can be performed using the two common methods: one-vs.-all and one-vs.-one [39]. The former method was applied in the LabVIEW program explicitly designed for classifying the hens' call.…”