Musical Genre Classification Using SVM and Audio Features

Abstract

The need of advance Music Information Retrieval increases as well as a huge amount of digital music files distribution on the internet. Musical genres are the main top-level descriptors used to organize digital music files. Most of work in labeling genre done manually. Thus, an automatic way for labeling a genre to digital music files is needed. The most standard approach to do automatic musical genre classification is feature extraction followed by supervised machine-learning. This research aims to find the best combination of audio features using several kernels of non-linear Support Vector Machines (SVM). The 31 different combinations of proposed audio features are dissimilar compared in any other related research. Furthermore, among the proposed audio features, Linear Predictive Coefficients (LPC) has not been used in another works related to musical genre classification. LPC was originally used for speech coding. An experimentation in classifying digital music file into a genre is carried out. The experiments are done by extracting feature sets related to timbre, rhythm, tonality and LPC from music files. All possible combination of the extracted features are classified using three different kernel of SVM classifier that are Radial Basis Function (RBF), polynomial and sigmoid. The result shows that the most appropriate kernel for automatic musical genre classification is polynomial kernel and the best combination of audio features is the combination of musical surface, Mel-Frequency Cepstrum Coefficients (MFCC), tonality and LPC. It achieves 76.6 % in classification accuracy.

Feature Extraction of Musical Instrument Tones using FFT and Segment Averaging

“…In the transform domain, there are two ways to get the feature extraction. The first one is based on fundamental signals [3][4][5][6] and the second one is not based on fundamental signals [7][8][9][10][11].…”

“…It still showing a relatively large number of the feature extraction coefficients. For example, based on the references, the feature extraction for musical instrument sounds gave 34 coefficients [7], 32 coefficients [8], 12 coefficients [9], 9 coefficients [10], and 8 coefficients [11]. Therefore, there is still an opportunity to further reduce the number of the feature extraction coefficients.…”

“…Table 6 compares the performance of several feature extraction for pianica tone recognition. As shown in Table 6, the feature extraction used in this research, when compared with the other feature extraction that have been published previously [7][8][9][10][11], is highly efficient for use in the musical instrument tones that have many significant local peaks in the Fourier transform domain. The reason is that it can give the number of feature extraction coefficients that are significantly much smaller.…”

“…Table 6. The comparison of the number of coefficients of the smallest feature extraction results, which resulted in a 100% recognition rate, of several feature extraction for pianica tone recognition Feature extraction Number of feature extraction coefficients Spectral Features [7] 34 DCT and windowing coefficients [8] 32 FFT and windowing coefficients [9] 12 Musical Surface [10] 9 DCT and segment averaging [11] 8 FFT and segment averaging (this research) 4…”

Feature Selection for Musical Genre Classification Using a Genetic Algorithm

“…Achieving optimal feature selection and extracting appropriate features and dimensions of the features reduction are very challenging tasks [16,17]. There are many different approaches that have been proposed to classify music genre, such as data mining [18,19], deep learning strategy [20,21,22,23,24], and machine learning stretegies such as hidden Markov model, AdaBoost, and support vector machine [25,26,27,28,29,30,31,32,33,34].…”