Research map / Label-efficient and robust learning / Recognition and long-tail
Open-World Deepfake Attribution via Confidence-Aware Asymmetric Learning
Haiyang Zheng, Nan Pu, Wenjing Li, Teng Long, Nicu Sebe, Zhun Zhong
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 L_ccr to base loss (ablation) | CAL · OW-DFA-40 Protocol-1 | All ACC: +18.4% All ACC, +32.8% Novel ACC | Improved |
| add L_acr on top of L_ccr (ablation) | CAL · OW-DFA-40 Protocol-1 | All ACC: +3.7% All ACC, +10.5% Novel ACC | Improved |
| add FFE (without L_acr) (ablation) | CAL · OW-DFA-40 Protocol-1 | All ACC: 80.1 -> 84.6 | Improved |
| add FFE (with L_acr) (ablation) | CAL · OW-DFA-40 Protocol-1 | All ACC: 87.4 -> 88.3 | Improved |
| use L_ccr instead of FixMatch threshold loss (ablation) | CAL · OW-DFA-40 Protocol-1 | Novel ACC: +1.5% All ACC, +4.5% Novel ACC | Improved |
| asymmetric thresholds instead of fixed gamma (ablation) | CAL · OW-DFA-40 Protocol-1 | Novel ACC: +2.4% All ACC, +3.6% Novel ACC | Improved |
| add Dynamic Prototype Pruning (estimate K_U) (ablation) | CAL · OW-DFA-40 Protocol-1 | All ACC: 87.5 -> 88.3 | Mixed |
| replace GCD class-number estimate with DPP | CAL · OW-DFA-40 Protocol-1 | class number error: 53.7 -> 4.8 | Improved |
| use CAL instead of CPL | CPL · OW-DFA-40 Protocol-1 | All ACC: 82.6 -> 88.3 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- ProtoGCD — ProtoGCD: Unified and Unbiased Prototype Learning for Generalized Category Discovery (TPAMI 2025) · better on 3 of 3 shared results
- OwMatch — OwMatch: Conditional Self-Labeling with Consistency for Open-World Semi-Supervised Learning (NeurIPS 2024) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- DF40: Toward Next-Generation Deepfake Detection (NeurIPS 2024 · similarity 0.88)
- TraceEvader: Making DeepFakes More Untraceable via Evading the Forgery Model Attribution (AAAI 2024 · similarity 0.88)
- From Specificity to Generality: Revisiting Generalizable Artifacts in Detecting Face Deepfakes (NeurIPS 2025 · similarity 0.87)
- Towards More General Video-based Deepfake Detection through Facial Component Guided Adaptation for Foundation Model (CVPR 2025 · similarity 0.87)
- FakeDiffer: Distributional Disparity Learning on Differentiated Reconstruction for Face Forgery Detection (AAAI 2025 · similarity 0.87)
- Veritas: Generalizable Deepfake Detection via Pattern-Aware Reasoning (ICLR 2026 · similarity 0.86)
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.
