Research map / Label-efficient and robust learning / Test-time adaptation
OST: Improving Generalization of DeepFake Detection via One-Shot Test-Time Training
Liang Chen, Yong Zhang, Yibing Song, Jue Wang, Lingqiao Liu
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 one-shot test-time training with pseudo-forgery (ablation) | OST detector · DFDC/DFD/DF1.0 (avg AUC) | AUC: 0.817 -> 0.854 | Improved |
| replace pseudo-forgery TTT with rotation-prediction TTT (ablation) | OST detector · DFDC/DFD/DF1.0 (avg AUC) | AUC: 0.817 -> 0.786 | Hurt |
| replace pseudo-forgery TTT with entropy minimization (TENT) (ablation) | OST detector · DFDC/DFD/DF1.0 (avg AUC) | AUC: 0.817 -> 0.819 | No clear effect |
| use more test-time gradient steps (1 to 3) (ablation) | OST · DFDC/DFD/DF1.0 (avg AUC) | AUC: 0.854 -> 0.861 | No clear effect |
| change template sampling from random to nearest neighbor or average (ablation) | OST · DFDC/DFD/DF1.0 (avg AUC) | AUC: 0.854 -> 0.855 (NN), 0.849 (Avg) | No clear effect |
| add OST to regular end-to-end Xception training (ablation) | Xception · DFDC/DFD/DF1.0 (avg AUC) | AUC: 0.820 -> 0.846 | Improved |
| use MAML meta-training instead of regular training (ablation) | OST · DFDC/DFD/DF1.0 (avg AUC) | AUC: 0.846 -> 0.854 | No clear effect |
| OST vs SRM, best prior detector, cross-dataset | OST · DFDC/DFD/DF1.0 (avg AUC) | AUC: 0.754 -> 0.854 | Improved |
| OST on in-dataset FF++ vs MADD | OST · FF++ test AUC | AUC: 0.998 -> 0.982 | Hurt |
Closest papers by meaning
- Exploring Unbiased Deepfake Detection via Token-Level Shuffling and Mixing (AAAI 2025 · similarity 0.89)
- D^3: Scaling Up Deepfake Detection by Learning from Discrepancy (CVPR 2025 · similarity 0.89)
- Exposing the Deception: Uncovering More Forgery Clues for Deepfake Detection (AAAI 2024 · similarity 0.89)
- Veritas: Generalizable Deepfake Detection via Pattern-Aware Reasoning (ICLR 2026 · similarity 0.89)
- DF40: Toward Next-Generation Deepfake Detection (NeurIPS 2024 · similarity 0.88)
- Transcending Forgery Specificity with Latent Space Augmentation for Generalizable Deepfake Detection (CVPR 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.
