Domain Generalization for Face Forgery Detection by Style Transfer
- Authors
- Kim, Taehoon; Choi, Jongwook; Cho, Hyunjin; Lim, Hyoungjun; Choi, Jongwon
- Issue Date
- Jan-2024
- Publisher
- Institute of Electrical and Electronics Engineers Inc.
- Keywords
- data augmentation; Deepfake detection; forgery detection; style transfer
- Citation
- Digest of Technical Papers - IEEE International Conference on Consumer Electronics, v.2024 IEEE
- Journal Title
- Digest of Technical Papers - IEEE International Conference on Consumer Electronics
- Volume
- 2024 IEEE
- URI
- https://scholarworks.bwise.kr/cau/handle/2019.sw.cau/73045
- DOI
- 10.1109/ICCE59016.2024.10444215
- ISSN
- 0747-668X
- Abstract
- Although deep fake detection models have made significant progress, the challenge of performance degradation remains yet for unseen datasets. To address this, we introduce a novel data generalization approach using style transfer to generate images in various domains. Utilizing style transfer, we create a new domain where domain-specific information is eliminated and subsequently train our model on the new domain. Our approach enhances the generalization performance of the detector by adding the style-transferred images to train the deepfake detector. Through the experiments, we confirm that the performance on the trained dataset remains unchanged while achieving an improvement of 8.8% on an unseen dataset. Therefore, We verify the effectiveness of the style-transferred images for generalizing the performance upon unseen datasets. © 2024 IEEE.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - Graduate School of Advanced Imaging Sciences, Multimedia and Film > Department of Imaging Science and Arts > 1. Journal Articles
![qrcode](https://api.qrserver.com/v1/create-qr-code/?size=55x55&data=https://scholarworks.bwise.kr/cau/handle/2019.sw.cau/73045)
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.