Research map / Video / Video segmentation and tracking
Adaptive Face Forgery Detection in Cross Domain
Luchuan Song, Zheng Fang, Xiaodan Li, Xiaoyi Dong, Zhenchao Jin, Yuefeng Chen, Siwei Lyu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add DICM (ablation) | Xception · FF++ c40 ablation | AUC: 0.925 -> 0.944 | Improved |
| add IDM (ablation) | Xception · FF++ c40 ablation | AUC: 0.925 -> 0.938 | Improved |
| add both DICM and IDM (ablation) | Xception · FF++ c40 ablation | AUC: 0.925 -> 0.952 | Improved |
| add frequency branch to DICM (ablation) | DICM · FF++ c40 ablation | AUC: 0.938 -> 0.944 | Improved |
| add cross-frame correlation to DICM (ablation) | DICM · FF++ c40 ablation | AUC: 0.932 -> 0.944 | Improved |
| use DICM instead of 3D conv for frame correlation (ablation) | Xception · FF++ c40 ablation | AUC: 0.931 -> 0.944 | Improved |
| use frequency instead of low-frequency or original image in DICM (ablation) | DICM · FF++ c40 ablation | AUC: 0.939 / 0.941 -> 0.944 | Improved |
| use IDM instead of normalized softmax (ablation) | Xception · FF++ c40 ablation | AUC: 0.924 -> 0.938 | Improved |
| use normalized softmax instead of softmax (ablation) | Xception · FF++ c40 ablation | AUC: 0.925 -> 0.924 | No clear effect |
| use IDM instead of ArcFace or SoftTriple (ablation) | Xception · FF++ c40 ablation | AUC: 0.923 / 0.926 -> 0.938 | Improved |
| apply bias to discriminative centers instead of embedding (ablation) | IDM · FF++ c40 ablation | AUC: 0.929 -> 0.938 | Improved |
| add DICM and IDM to reduce prediction instability (ablation) | Xception · FF++ c40 ablation | PUP0.7: 66.1% -> 26.1% | Improved |
Closest papers by meaning
- Rethinking Vision-Language Model in Face Forensics: Multi-Modal Interpretable Forged Face Detector (CVPR 2025 · similarity 0.90)
- FreqBlender: Enhancing DeepFake Detection by Blending Frequency Knowledge (NeurIPS 2024 · similarity 0.90)
- FakeDiffer: Distributional Disparity Learning on Differentiated Reconstruction for Face Forgery Detection (AAAI 2025 · similarity 0.88)
- Identity-Aware Vision-Language Model for Explainable Face Forgery Detection (AAAI 2026 · similarity 0.88)
- DevFD : Developmental Face Forgery Detection by Learning Shared and Orthogonal LoRA Subspaces (NeurIPS 2025 · similarity 0.88)
- DF40: Toward Next-Generation Deepfake Detection (NeurIPS 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
