A Hybrid CNN and RNN Variant Model for Music Classification
Music genre classification has a significant role in information retrieval for the organization of growing collections of music. It is challenging to classify music with reliable accuracy. Many methods have utilized handcrafted features to identify unique patterns but are still unable to determine t...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-01-01
|
Series: | Applied Sciences |
Subjects: | |
Online Access: | https://www.mdpi.com/2076-3417/13/3/1476 |
_version_ | 1797625205954183168 |
---|---|
author | Mohsin Ashraf Fazeel Abid Ikram Ud Din Jawad Rasheed Mirsat Yesiltepe Sook Fern Yeo Merve T. Ersoy |
author_facet | Mohsin Ashraf Fazeel Abid Ikram Ud Din Jawad Rasheed Mirsat Yesiltepe Sook Fern Yeo Merve T. Ersoy |
author_sort | Mohsin Ashraf |
collection | DOAJ |
description | Music genre classification has a significant role in information retrieval for the organization of growing collections of music. It is challenging to classify music with reliable accuracy. Many methods have utilized handcrafted features to identify unique patterns but are still unable to determine the original music characteristics. Comparatively, music classification using deep learning models has been shown to be dynamic and effective. Among the many neural networks, the combination of a convolutional neural network (CNN) and variants of a recurrent neural network (RNN) has not been significantly considered. Additionally, addressing the flaws in the particular neural network classification model, this paper proposes a hybrid architecture of CNN and variants of RNN such as long short-term memory (LSTM), Bi-LSTM, gated recurrent unit (GRU), and Bi-GRU. We also compared the performance based on Mel-spectrogram and Mel-frequency cepstral coefficient (MFCC) features. Empirically, the proposed hybrid architecture of CNN and Bi-GRU using Mel-spectrogram achieved the best accuracy at 89.30%, whereas the hybridization of CNN and LSTM using MFCC achieved the best accuracy at 76.40%. |
first_indexed | 2024-03-11T09:53:16Z |
format | Article |
id | doaj.art-b607c8cc97ac4253968ca40f36c15775 |
institution | Directory Open Access Journal |
issn | 2076-3417 |
language | English |
last_indexed | 2024-03-11T09:53:16Z |
publishDate | 2023-01-01 |
publisher | MDPI AG |
record_format | Article |
series | Applied Sciences |
spelling | doaj.art-b607c8cc97ac4253968ca40f36c157752023-11-16T16:06:01ZengMDPI AGApplied Sciences2076-34172023-01-01133147610.3390/app13031476A Hybrid CNN and RNN Variant Model for Music ClassificationMohsin Ashraf0Fazeel Abid1Ikram Ud Din2Jawad Rasheed3Mirsat Yesiltepe4Sook Fern Yeo5Merve T. Ersoy6Department of Computer Science, University of Central Punjab, Lahore 54700, PakistanDepartment of Information Systems, University of Management and Technology, Lahore 54700, PakistanDepartment of Information Technology, University of Haripur, Haripur 22610, PakistanDepartment of Software Engineering, Nisantasi University, Istanbul 34398, TurkeyDepartment of Mathematical Engineering, Yildiz Technical University, Istanbul 34220, TurkeyFaculty of Business, Multimedia University, Melaka 75450, MalaysiaDepartment of Software Engineering, Nisantasi University, Istanbul 34398, TurkeyMusic genre classification has a significant role in information retrieval for the organization of growing collections of music. It is challenging to classify music with reliable accuracy. Many methods have utilized handcrafted features to identify unique patterns but are still unable to determine the original music characteristics. Comparatively, music classification using deep learning models has been shown to be dynamic and effective. Among the many neural networks, the combination of a convolutional neural network (CNN) and variants of a recurrent neural network (RNN) has not been significantly considered. Additionally, addressing the flaws in the particular neural network classification model, this paper proposes a hybrid architecture of CNN and variants of RNN such as long short-term memory (LSTM), Bi-LSTM, gated recurrent unit (GRU), and Bi-GRU. We also compared the performance based on Mel-spectrogram and Mel-frequency cepstral coefficient (MFCC) features. Empirically, the proposed hybrid architecture of CNN and Bi-GRU using Mel-spectrogram achieved the best accuracy at 89.30%, whereas the hybridization of CNN and LSTM using MFCC achieved the best accuracy at 76.40%.https://www.mdpi.com/2076-3417/13/3/1476music classificationmusic information retrievalconvolutional neural networkrecurrent neural networkMel-spectrogram |
spellingShingle | Mohsin Ashraf Fazeel Abid Ikram Ud Din Jawad Rasheed Mirsat Yesiltepe Sook Fern Yeo Merve T. Ersoy A Hybrid CNN and RNN Variant Model for Music Classification Applied Sciences music classification music information retrieval convolutional neural network recurrent neural network Mel-spectrogram |
title | A Hybrid CNN and RNN Variant Model for Music Classification |
title_full | A Hybrid CNN and RNN Variant Model for Music Classification |
title_fullStr | A Hybrid CNN and RNN Variant Model for Music Classification |
title_full_unstemmed | A Hybrid CNN and RNN Variant Model for Music Classification |
title_short | A Hybrid CNN and RNN Variant Model for Music Classification |
title_sort | hybrid cnn and rnn variant model for music classification |
topic | music classification music information retrieval convolutional neural network recurrent neural network Mel-spectrogram |
url | https://www.mdpi.com/2076-3417/13/3/1476 |
work_keys_str_mv | AT mohsinashraf ahybridcnnandrnnvariantmodelformusicclassification AT fazeelabid ahybridcnnandrnnvariantmodelformusicclassification AT ikramuddin ahybridcnnandrnnvariantmodelformusicclassification AT jawadrasheed ahybridcnnandrnnvariantmodelformusicclassification AT mirsatyesiltepe ahybridcnnandrnnvariantmodelformusicclassification AT sookfernyeo ahybridcnnandrnnvariantmodelformusicclassification AT mervetersoy ahybridcnnandrnnvariantmodelformusicclassification AT mohsinashraf hybridcnnandrnnvariantmodelformusicclassification AT fazeelabid hybridcnnandrnnvariantmodelformusicclassification AT ikramuddin hybridcnnandrnnvariantmodelformusicclassification AT jawadrasheed hybridcnnandrnnvariantmodelformusicclassification AT mirsatyesiltepe hybridcnnandrnnvariantmodelformusicclassification AT sookfernyeo hybridcnnandrnnvariantmodelformusicclassification AT mervetersoy hybridcnnandrnnvariantmodelformusicclassification |