An improved deepfake detection method based on CNNS
Today's image generation technology can generate high-quality face images, and it isn't easy to recognize the authenticity of the generated images through human eyes. This study aims to improve deepfake detection, a face swapping forgery, by absorbing the advantages of deep learning techno...
Saved in:
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Little Lion Scientific Islamabad Pakistan
2022
|
Online Access: | http://eprints.utem.edu.my/id/eprint/27028/2/32Vol100No17.pdf http://eprints.utem.edu.my/id/eprint/27028/ https://www.jatit.org/volumes/Vol100No17/32Vol100No17.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
id |
my.utem.eprints.27028 |
---|---|
record_format |
eprints |
spelling |
my.utem.eprints.270282024-01-16T10:35:28Z http://eprints.utem.edu.my/id/eprint/27028/ An improved deepfake detection method based on CNNS Dafeng, Gong Jaya Kumar, Yogan Goh, Ong Sing Ye Zi Choo, Yun Huoy Wanle, Chi Today's image generation technology can generate high-quality face images, and it isn't easy to recognize the authenticity of the generated images through human eyes. This study aims to improve deepfake detection, a face swapping forgery, by absorbing the advantages of deep learning technologies. This study generates a unified and enhanced data set from multiple sources using spatial enhancement technology to solve the problem of poor detection performance on cross-data sets. Taking the advantages of Inception and ResNet networks, new deepfake detection architecture composed of 20 network layers is proposed as the deepfake detection model. To further improve the proposed model, hyperparameter values are optimized. The experiment result shows that the proposed network significantly enhanced over the mainstream methods, such as ResNeXt50, ResNet101, XceptionNet, and VGG19, in terms of accuracy, loss value, AUC, numbers of parameters, and FLOPs. Overall, the methods introduced in this study can help to expand the data set, better detect deepfake contents, and effectively optimize network models Little Lion Scientific Islamabad Pakistan 2022-09 Article PeerReviewed text en http://eprints.utem.edu.my/id/eprint/27028/2/32Vol100No17.pdf Dafeng, Gong and Jaya Kumar, Yogan and Goh, Ong Sing and Ye Zi and Choo, Yun Huoy and Wanle, Chi (2022) An improved deepfake detection method based on CNNS. Journal of Theoretical and Applied Information Technology, 100 (17). 5684 - 5691. ISSN 1992-8645 https://www.jatit.org/volumes/Vol100No17/32Vol100No17.pdf |
institution |
Universiti Teknikal Malaysia Melaka |
building |
UTEM Library |
collection |
Institutional Repository |
continent |
Asia |
country |
Malaysia |
content_provider |
Universiti Teknikal Malaysia Melaka |
content_source |
UTEM Institutional Repository |
url_provider |
http://eprints.utem.edu.my/ |
language |
English |
description |
Today's image generation technology can generate high-quality face images, and it isn't easy to recognize the authenticity of the generated images through human eyes. This study aims to improve deepfake detection, a face swapping forgery, by absorbing the advantages of deep learning technologies. This study generates a unified and enhanced data set from multiple sources using spatial enhancement technology to solve the problem of poor detection performance on cross-data sets. Taking the advantages of Inception and ResNet networks, new deepfake detection architecture composed of 20 network layers is proposed as the deepfake detection model. To further improve the proposed model, hyperparameter values are optimized. The experiment result shows that the proposed network significantly enhanced over the mainstream methods, such as ResNeXt50, ResNet101, XceptionNet, and VGG19, in terms of accuracy, loss value, AUC, numbers of parameters, and FLOPs. Overall, the methods introduced in this study can help to expand the data set, better detect deepfake contents, and effectively optimize network models |
format |
Article |
author |
Dafeng, Gong Jaya Kumar, Yogan Goh, Ong Sing Ye Zi Choo, Yun Huoy Wanle, Chi |
spellingShingle |
Dafeng, Gong Jaya Kumar, Yogan Goh, Ong Sing Ye Zi Choo, Yun Huoy Wanle, Chi An improved deepfake detection method based on CNNS |
author_facet |
Dafeng, Gong Jaya Kumar, Yogan Goh, Ong Sing Ye Zi Choo, Yun Huoy Wanle, Chi |
author_sort |
Dafeng, Gong |
title |
An improved deepfake detection method based on CNNS |
title_short |
An improved deepfake detection method based on CNNS |
title_full |
An improved deepfake detection method based on CNNS |
title_fullStr |
An improved deepfake detection method based on CNNS |
title_full_unstemmed |
An improved deepfake detection method based on CNNS |
title_sort |
improved deepfake detection method based on cnns |
publisher |
Little Lion Scientific Islamabad Pakistan |
publishDate |
2022 |
url |
http://eprints.utem.edu.my/id/eprint/27028/2/32Vol100No17.pdf http://eprints.utem.edu.my/id/eprint/27028/ https://www.jatit.org/volumes/Vol100No17/32Vol100No17.pdf |
_version_ |
1789429988088348672 |
score |
13.18916 |