Audio-Based Music Classification with DenseNet and Data Augmentation

Abstract

In recent years, deep learning technique has received intense attention owing to its great success in image recognition. A tendency of adaption of deep learning in various information processing fields has formed, including music information retrieval (MIR). In this paper, we conduct a comprehensive study on music audio classification with improved convolutional neural networks (CNNs). To the best of our knowledge, this the first work to apply Densely Connected Convolutional Networks (DenseNet) to music audio tagging, which has been demonstrated to perform better than Residual neural network (ResNet). Additionally, two specific data augmentation approaches of time overlapping and pitch shifting have been proposed to address the deficiency of labelled data in the MIR. Moreover, an ensemble learning of stacking is employed based on SVM. We believe that the proposed combination of strong representation of DenseNet and data augmentation can be adapted to other audio processing tasks.

Combining Reduction and Dense Blocks for Music Genre Classification

“…It was able to achieve an accuracy of 93.4% through 10-fold crossvalidation over the GTZAN dataset. Several approaches used CNN-based networks but were not able to exceed the accuracy of 91% such as [1,11,12,13]. Others tried refining their results by overcoming the blurry classification of certain genres inside the GTZAN dataset.…”

Neural Models for Target-Based Computer-Assisted Musical Orchestration: A Preliminary Study

“…Thus, when an unknown target sound is given as input, the network will identify which are the best instruments to match the target sound, and it will be able to deconstruct a complex mixture of timbres into individual instrument notes. This method is motivated by the good results obtained in previous research on musical instruments identification (Benetos, Kotti, & Kotropoulos, 2007;Kitahara, Goto, & Okuno, 2005) and the more recent use of deep neural networks for musical classification (Lostanlen & Cella, 2016;Bian et al, 2019).…”

MS-SincResNet: Joint Learning of 1D and 2D Kernels Using Multi-scale SincNet and ResNet for Music Genre Classification

“…In recent years, with the remarkable success of deep learning techniques in computer vision applications, deep neural networks (DNNs) have also shown great success in speech/music classification or recognition tasks, such as speaker recognition [36,43], music genre classification [6,39], speech emotion recognition [49], etc. In these tasks, deep learning provides a new way to extract discriminative embeddings from those famous hand-crafted acoustic features, called i-vector content, for classification/recognition [8].…”

“…In these tasks, deep learning provides a new way to extract discriminative embeddings from those famous hand-crafted acoustic features, called i-vector content, for classification/recognition purposes. To this end, deep learning methods based on convolutional neural networks (CNNs) are the most widely used approach to obtain embeddings from those i-vector content, such as MFCC [46,47,53], OSC coefficients [54], 2D representations like audio spectrogram or chromagram [6,39], etc. Bisharad et al proposed a music genre classification system using residual neural network (ResNet) based model [8].…”

“…purposes. To this end, deep learning methods based on convolutional neural networks (CNNs) are the most widely used approach to obtain embeddings from those i-vector content, such as MFCC[46,47,53], OSC coefficients[54], 2D representations like audio spectrogram or chromagram[6,39], etc.Bisharad et al proposed a music genre classification system using residual neural network (ResNet) based model…”