Ernie Chu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Ernie Chu has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
- Jun-Cheng Chen
Academia Sinica, Taipei, Taiwan
2 shared publications - Kai-Lung Hua
National Taiwan University of Science and Technology
2 shared publications - Yue-Hua Han
Academia Sinica
2 shared publications - Po-Han Huang
National Taiwan University of Science and Technology, TW
1 shared publication - Shu-Tzu Lo
National Taiwan University of Science and Technology
1 shared publication - Tai-Ming Huang
National Taiwan University, Taipei, Taiwan and Academia Sinica, Taipei, Taiwan
1 shared publication
Latest Publications
- 4 December 2024
- Published by Springer Nature
- p. 194-210
Generalized Image-Based Deepfake Detection Through Foundation Model Adaptation
- 4 December 2024
- Published by Springer Nature
- p. 194-210
Generalized Image-Based Deepfake Detection Through Foundation Model Adaptation
- Open Access
- 6 December 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Multi-Task Self-Blended Images for Face Forgery Detection
Deepfake detection has attracted extensive attention due to widespread forged images on social media. Recently, self-supervised learning (SSL) based Deepfake detection approaches have outperformed supervised methods in terms of model generalization. However, we notice that most SSL-based methods do not take the manipulation strength levels of synthesized forgery samples into consideration according to different synthesis parameters and result in suboptimal detection performances. To address this issue, we introduce several auxiliary losses to the state-of-the-art SSL-based method based on different synthesis sub-tasks during data generation by inferring their synthesis parameters where the ground-truth labels are obtained from the synthesis pipeline for free. With comprehensive evaluations on various benchmarks, our approach has achieved noticeable performance improvement. Specifically, for the cross-dataset evaluation, the proposed approach outperforms the state-of-the-art method in terms of AUC on various datasets with improvements of 3.4%, 1.47%, 1.56%, and 1.3% on the CDF, DFDC, DFDCP, and FFIW datasets and achieves competitive performance on the DFD dataset. This further demonstrates the effectiveness of the proposed approach in its generalization ability.
- 6 December 2023
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessMulti-Task Self-Blended Images for Face Forgery Detection
Deepfake detection has attracted extensive attention due to widespread forged images on social media. Recently, self-supervised learning (SSL) based Deepfake detection approaches have outperformed supervised methods in terms of model generalization. However, we notice that most SSL-based methods do not take the manipulation strength levels of synthesized forgery samples into consideration according to different synthesis parameters and result in suboptimal detection performances. To address this issue, we introduce several auxiliary losses to the state-of-the-art SSL-based method based on different synthesis sub-tasks during data generation by inferring their synthesis parameters where the ground-truth labels are obtained from the synthesis pipeline for free. With comprehensive evaluations on various benchmarks, our approach has achieved noticeable performance improvement. Specifically, for the cross-dataset evaluation, the proposed approach outperforms the state-of-the-art method in terms of AUC on various datasets with improvements of 3.4%, 1.47%, 1.56%, and 1.3% on the CDF, DFDC, DFDCP, and FFIW datasets and achieves competitive performance on the DFD dataset. This further demonstrates the effectiveness of the proposed approach in its generalization ability.