Music Genre Classification using Spectral Analysis Techniques With Hybrid Convolution-Recurrent Neural Network

Abstract

In this work, the objective is to classify the audio data into specific genres from GTZAN dataset which contain about 10 genres. First, it perform the audio splitting to make it signal into clips which contains homogeneous content. Short-term Fourier Transform (STFT), Mel-spectrogram and Mel-frequency cepstrum coefficient (MFCC) are the most common feature extraction technique and each feature extraction technique has been successful in their own various audio applications. Then, these feature extractions of the audio fed to the Convolution Neural Network (CNN) model and VGG16 Neural Network model, which consist of 16 convolution layers network. We perform different feature extraction with different CNN and VGG16 model with or without different Recurrent Neural Network (RNN) and evaluated performance measure. In this model, it has achieved overall accuracy 95.5\% for this task.

Hyperparameter Optimization of CNN Classifier for Music Genre Classification

“…The time duration of the data affects the accuracy of results obtained, as seen in the comparison of several studies in the duration data comparison (Table 6). Research Time duration Acc % Ahmad [7] 3 seconds 95 Lau [19] 3 seconds 81.73 Zhang [10] 3 seconds 87,4 Vita [9] 30 seconds 58 Purnama [13] 30 seconds 60 Ndou [3] 30…”

“…These features can be used to classify the type of music. The research conducted by Shah et al [6] [7]. In another study, classification using CNN with a three-second music duration feature gave 72.4% better accuracy than a thirty-second music duration feature which was only 53.50%; the spectrogram feature showed increased accuracy but with an even greater number of epochs [3].…”

A Low Cost Solution for Automatic Plastic Segregation

“…[3] At present, there are few popular techniques to sort plastic and non-plastic materials. One of them is infrared hyper spectral imaging which is used for real-time instantaneous identification of plastics, where intelligent algorithms for image recognition and classification are used [6][7][8][9][10][11]. Near Infrared Spectroscopy is another technique used for the identification and selection of materials and is used for plastic segregation as proposed in [14].…”

“…Audio signals from different materials are transformed to data sets that represents the material explicitly using feature engineering. [4][5][6][7] There are many feature extraction techniques for audio signal [19] and these feature extraction methods use both spectral, and joint time-frequency signal representation. Artificial Neural Network is the classifier used to classify the feature database.…”