GFADE: generalized feature adaptation and discrimination enhancement for deepfake detection
With the rapid advancement of deep generative techniques, such as generative adversarial networks (GANs), the creation of realistic fake images and videos has become increasingly accessible, raising significant security and privacy concerns. Although existing deepfake detection methods perform well...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
PeerJ Inc.
2025-05-01
|
| Series: | PeerJ Computer Science |
| Subjects: | |
| Online Access: | https://peerj.com/articles/cs-2879.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850279513809747968 |
|---|---|
| author | ZhiYong Tian Junkai Yi |
| author_facet | ZhiYong Tian Junkai Yi |
| author_sort | ZhiYong Tian |
| collection | DOAJ |
| description | With the rapid advancement of deep generative techniques, such as generative adversarial networks (GANs), the creation of realistic fake images and videos has become increasingly accessible, raising significant security and privacy concerns. Although existing deepfake detection methods perform well within a single dataset, they often experience substantial performance degradation when applied across datasets or manipulation types. To address this challenge, we propose a novel deepfake detection framework that combines multiple loss functions and the MixStyle technique. By integrating Cross-Entropy Loss, ArcFace loss, and Focal Loss, our model enhances its discriminative power to better handle complex forgery characteristics and effectively mitigate data imbalance. Additionally, the MixStyle technique introduces diverse visual styles during training, further improving the model’s generalization across different datasets and manipulation scenarios. Experimental results demonstrate that our method achieves superior detection accuracy across a range of cross-dataset and cross-manipulation tests, significantly improving model robustness and generalizability. |
| format | Article |
| id | doaj-art-36a8dc0f00d54c0b8a5d8d8f29ea89ec |
| institution | OA Journals |
| issn | 2376-5992 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | PeerJ Inc. |
| record_format | Article |
| series | PeerJ Computer Science |
| spelling | doaj-art-36a8dc0f00d54c0b8a5d8d8f29ea89ec2025-08-20T01:49:04ZengPeerJ Inc.PeerJ Computer Science2376-59922025-05-0111e287910.7717/peerj-cs.2879GFADE: generalized feature adaptation and discrimination enhancement for deepfake detectionZhiYong Tian0Junkai Yi1School of Automation, Beijing Information Science and Technology University, Beijing, ChinaKey Laboratory of Modern Measurement and Control Technology Ministry of Education, Beijing Information Science and Technology University, Beijing, ChinaWith the rapid advancement of deep generative techniques, such as generative adversarial networks (GANs), the creation of realistic fake images and videos has become increasingly accessible, raising significant security and privacy concerns. Although existing deepfake detection methods perform well within a single dataset, they often experience substantial performance degradation when applied across datasets or manipulation types. To address this challenge, we propose a novel deepfake detection framework that combines multiple loss functions and the MixStyle technique. By integrating Cross-Entropy Loss, ArcFace loss, and Focal Loss, our model enhances its discriminative power to better handle complex forgery characteristics and effectively mitigate data imbalance. Additionally, the MixStyle technique introduces diverse visual styles during training, further improving the model’s generalization across different datasets and manipulation scenarios. Experimental results demonstrate that our method achieves superior detection accuracy across a range of cross-dataset and cross-manipulation tests, significantly improving model robustness and generalizability.https://peerj.com/articles/cs-2879.pdfDeepfake detectionMixStyleArcFace lossFocal loss |
| spellingShingle | ZhiYong Tian Junkai Yi GFADE: generalized feature adaptation and discrimination enhancement for deepfake detection PeerJ Computer Science Deepfake detection MixStyle ArcFace loss Focal loss |
| title | GFADE: generalized feature adaptation and discrimination enhancement for deepfake detection |
| title_full | GFADE: generalized feature adaptation and discrimination enhancement for deepfake detection |
| title_fullStr | GFADE: generalized feature adaptation and discrimination enhancement for deepfake detection |
| title_full_unstemmed | GFADE: generalized feature adaptation and discrimination enhancement for deepfake detection |
| title_short | GFADE: generalized feature adaptation and discrimination enhancement for deepfake detection |
| title_sort | gfade generalized feature adaptation and discrimination enhancement for deepfake detection |
| topic | Deepfake detection MixStyle ArcFace loss Focal loss |
| url | https://peerj.com/articles/cs-2879.pdf |
| work_keys_str_mv | AT zhiyongtian gfadegeneralizedfeatureadaptationanddiscriminationenhancementfordeepfakedetection AT junkaiyi gfadegeneralizedfeatureadaptationanddiscriminationenhancementfordeepfakedetection |