IMPLEMENTASI CONVOLUTIONAL NEURAL NETWORK (CNN) DENGAN EKSTRAKSI FITUR MFCC DAN CHROMA FEATURES DALAM KLASIFIKASI GENRE MUSIK

HAKIM, ELAN ABDUL and Utami, Alvi Syahrini (2025) IMPLEMENTASI CONVOLUTIONAL NEURAL NETWORK (CNN) DENGAN EKSTRAKSI FITUR MFCC DAN CHROMA FEATURES DALAM KLASIFIKASI GENRE MUSIK. Undergraduate thesis, Sriwijaya University.

[thumbnail of RAMA_55201_09021182126030_cover.jpg] Image
RAMA_55201_09021182126030_cover.jpg - Accepted Version
Available under License Creative Commons Public Domain Dedication.

Download (1MB)
[thumbnail of RAMA_55201_09021182126030.pdf] Text
RAMA_55201_09021182126030.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (2MB) | Request a copy
[thumbnail of RAMA_55201_09021182126030_TURNITIN.pdf] Text
RAMA_55201_09021182126030_TURNITIN.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (5MB) | Request a copy
[thumbnail of RAMA_55201_09021182126030_0022127804_01_front_ref.pdf] Text
RAMA_55201_09021182126030_0022127804_01_front_ref.pdf - Accepted Version
Available under License Creative Commons Public Domain Dedication.

Download (1MB)
[thumbnail of RAMA_55201_09021182126030_0022127804_02.pdf] Text
RAMA_55201_09021182126030_0022127804_02.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (457kB) | Request a copy
[thumbnail of RAMA_55201_09021182126030_0022127804_03.pdf] Text
RAMA_55201_09021182126030_0022127804_03.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (321kB) | Request a copy
[thumbnail of RAMA_55201_09021182126030_0022127804_04.pdf] Text
RAMA_55201_09021182126030_0022127804_04.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (876kB) | Request a copy
[thumbnail of RAMA_55201_09021182126030_0022127804_05.pdf] Text
RAMA_55201_09021182126030_0022127804_05.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (548kB) | Request a copy
[thumbnail of RAMA_55201_09021182126030_0022127804_06.pdf] Text
RAMA_55201_09021182126030_0022127804_06.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (241kB) | Request a copy
[thumbnail of RAMA_55201_09021182126030_0022127804_07_ref.pdf] Text
RAMA_55201_09021182126030_0022127804_07_ref.pdf - Bibliography
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (242kB) | Request a copy
[thumbnail of RAMA_55201_09021182126030_0022127804_08_lamp.pdf] Text
RAMA_55201_09021182126030_0022127804_08_lamp.pdf - Accepted Version
Restricted to Repository staff only
Available under License Creative Commons Public Domain Dedication.

Download (474kB) | Request a copy

Abstract

Music is an important part of human life and continues to evolve with the advancement of information technology. The diversity of music genres that emerge due to differences in instruments, rhythm, technique, and lyrics presents a unique challenge in accurately classifying genres. This study aims to develop a music genre classification system using the Convolutional Neural Network (CNN) method with a Resnet-50 architecture and additional extraction of MFCC and Chroma Features, as well as to evaluate the accuracy achieved by various model configurations. The data used consists of 30-second .wav audio files from the GTZAN dataset. This dataset contains 1,000 audio files divided into 10 genre classes: blues, classical, country, disco, hip hop, jazz, metal, pop, reggae, and rock. Audio samples are extracted for their MFCC and Chroma features, which are then represented as 2D arrays that can be visualized as spectrograms and processed by the CNN model. The results of the study show that the CNN with the Resnet-50 architecture achieved fairly good performance, with an accuracy of 66%, precision of 68.86%, recall of 66%, and F1 score of 65.86%.

Item Type: Thesis (Undergraduate)
Uncontrolled Keywords: Convolutional Neural Network (CNN), Resnet-50, klasifikasi genre musik, GTZAN, MFCC, Chroma Features
Subjects: T Technology > T Technology (General) > T1-995 Technology (General)
Divisions: 09-Faculty of Computer Science > 55201-Informatics (S1)
Depositing User: Elan Abdul Hakim
Date Deposited: 30 Jun 2025 07:38
Last Modified: 30 Jun 2025 07:38
URI: http://repository.unsri.ac.id/id/eprint/176161

Actions (login required)

View Item View Item