2 papers
cs.CV2025
Contrastive Desensitization Learning for Cross Domain Face Forgery Detection
Lingyu Qiu, Ke Jiang, Xiaoyang Tan
In this paper, we propose a new cross-domain face forgery detection method that is insensitive to different and possibly unseen forgery methods while ensuring an acceptable low fal…
cs.CV2025
RoGA: Towards Generalizable Deepfake Detection through Robust Gradient Alignment
Lingyu Qiu, Ke Jiang, Xiaoyang Tan
Recent advancements in domain generalization for deepfake detection have attracted significant attention, with previous methods often incorporating additional modules to prevent ov…