Research map / Label-efficient and robust learning / Recognition and long-tail
Dual Contrastive Learning for General Face Forgery Detection
Ke Sun, Taiping Yao, Shen Chen, Shouhong Ding, Jilin Li, Rongrong Ji
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 Inter-ICL, special data views, hard sample selection, and Intra-ICL (ablation) | Face forgery detection · Celeb-DF | AUC: 80.30 -> 82.30 | Improved |
| Add Inter-ICL, special data views, hard sample selection, and Intra-ICL (ablation) | Face forgery detection · DFD | AUC: 90.12 -> 91.66 | Improved |
| Add specially designed data views to Inter-ICL (ablation) | DCL · Celeb-DF | AUC: 76.81 -> 79.34 | Improved |
| Add specially designed data views to Inter-ICL (ablation) | DCL · DFD | AUC: 88.03 -> 89.24 | Improved |
| Add hard sample selection to Inter-ICL with special data views (ablation) | DCL · Celeb-DF | AUC: 79.34 -> 80.30 | Improved |
| Add hard sample selection to Inter-ICL with special data views (ablation) | DCL · DFD | AUC: 89.24 -> 90.12 | Improved |
| Use DCL rather than Local-relation | Cross-database face forgery detection · FF++ (HQ), intra-domain | AUC: 99.46 -> 99.30 | Hurt |
| Use DCL rather than Local-relation | Cross-database face forgery detection · DFD | AUC: 89.24 -> 91.66 | Improved |
| Use DCL rather than Local-relation | Cross-database face forgery detection · DFDC | AUC: 76.53 -> 76.71 | Improved |
| Use DCL rather than MAT (EN-b4) | Cross-database face forgery detection · Wild Deepfake | AUC: 70.15 -> 71.14 | Improved |
| Use DCL rather than Local-relation | Cross-database face forgery detection · Celeb-DF | AUC: 78.26 -> 82.30 | Improved |
| Use DCL rather than SPSL | Cross-dataset evaluation from FF++ (LQ) · Celeb-DF | AUC: 76.88 -> 81.00 | Improved |
Closest papers by meaning
- Frequency-Aware Discriminative Feature Learning Supervised by Single-Center Loss for Face Forgery Detection (CVPR 2021 · similarity 0.90)
- Rethinking Vision-Language Model in Face Forensics: Multi-Modal Interpretable Forged Face Detector (CVPR 2025 · similarity 0.88)
- Stacking Brick by Brick: Aligned Feature Isolation for Incremental Face Forgery Detection (CVPR 2025 · similarity 0.88)
- A Hitchhiker's Guide to Fine-Grained Face Forgery Detection Using Common Sense Reasoning (NeurIPS 2024 · 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 0.87)
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.
