Research map / Label-efficient and robust learning / Domain adaptation and generalization
Generative Domain Adaptation for Face Anti-Spoofing
Qianyu Zhou, Ke-Yue Zhang, Taiping Yao, Ran Yi, Kekai Sheng, Shouhong Ding, Lizhuang Ma
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 neural statistic consistency (NSC) (ablation) | source-trained FAS model · I&C&M to O | AUC (%): 80.9 -> 84.9 | Improved |
| add NSC (ablation) | source-trained FAS model · O&C&I to M | HTER (%): 29.2 -> 20.0 | Improved |
| add NSC (ablation) | source-trained FAS model · O&M&I to C | HTER (%): 32.7 -> 28.7 | Improved |
| add dual-level semantic consistency (DSC) on top of NSC (ablation) | GDA · O&M&I to C | HTER (%): 28.7 -> 14.4 | Improved |
| add DSC on top of NSC (ablation) | GDA · O&C&I to M | AUC (%): 89.0 -> 92.0 | Improved |
| add DSC on top of NSC (ablation) | GDA · I&C&M to O | HTER (%): 21.1 -> 16.5 | Improved |
| add intra-domain spectrum mixup (SpecMix) (ablation) | GDA with NSC+DSC · O&C&I to M | HTER (%): 14.1 -> 9.2 | Improved |
| add SpecMix (ablation) | GDA with NSC+DSC · O&M&I to C | HTER (%): 14.4 -> 12.2 | Improved |
| add SpecMix (ablation) | GDA with NSC+DSC · O&C&M to I | AUC (%): 91.5 -> 96.0 | Improved |
| add SpecMix (ablation) | GDA with NSC+DSC · I&C&M to O | HTER (%): 16.5 -> 14.4 | Improved |
| raise SpecMix eta from 0.1 to 0.5 during training (ablation) | GDA · O&C&I to M | AUC (%): 98.0 -> 97.2 | No clear effect |
| raise SpecMix eta from 0.1 to 0.4 at inference (ablation) | GDA on Idiap Replay-Attack · Replay-Attack | HTER (%): 0 -> 0 | No clear effect |
Closest papers by meaning
- Source-Free Domain Adaptation With Domain Generalized Pretraining for Face Anti-Spoofing (TPAMI 2024 · similarity 0.93)
- Source-Free Domain Adaptation with Contrastive Domain Alignment and Self-Supervised Exploration for Face Anti-Spoofing (ECCV 2022 · similarity 0.93)
- Gradient Alignment for Cross-Domain Face Anti-Spoofing (CVPR 2024 · similarity 0.91)
- Test-Time Domain Generalization for Face Anti-Spoofing (CVPR 2024 · similarity 0.90)
- Domain-Hallucinated Updating for Multi-Domain Face Anti-spoofing (AAAI 2024 · similarity 0.90)
- CFPL-FAS: Class Free Prompt Learning for Generalizable Face Anti-spoofing (CVPR 2024 · similarity 0.89)
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.
