Research map / Label-efficient and robust learning / Knowledge distillation
Meta-Teacher For Face Anti-Spoofing
Yunxiao Qin, Zitong Yu, Longbin Yan, Zezheng Wang, Chenxu Zhao, Zhen Lei
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 |
|---|---|---|---|
| supervise detector with meta-teacher instead of facial depth label (FAS-DR) | FAS-DR · OULU-NPU Protocol 1 | ACER: 1.5 -> 0.6 | Improved |
| supervise CDCN with meta-teacher instead of depth label | CDCN · OULU-NPU Protocol 4 | ACER: 6.9 -> 3.7 | Improved |
| supervise DTN with meta-teacher instead of pixel-wise binary mask label | DTN · SiW-M average | ACER: 16.8 -> 13.1 | Improved |
| supervise FAS-DR with meta-teacher instead of depth label | FAS-DR · SiW-M average | ACER: 12.6 -> 10.5 | Improved |
| supervise FAS-DR-BC with meta-teacher instead of depth label | FAS-DR-BC · Domain generalization benchmark | HTER: 19.14 -> 11.93 (O&C&M to I) | Improved |
| supervise RFMetaFAS with meta-teacher instead of facial depth | RFMetaFAS* · Domain generalization, O&C&M to I | HTER: 18.89 -> 12.77 | Mixed |
| use meta-teacher instead of Distill/BANs/Fitnet/OKDDip teachers | FAS-DR · OULU-NPU Protocol 4 | ACER: 5.5 -> 4.3 | Improved |
| add pretraining of meta-teacher before bi-level optimization (ablation) | MT-FAS · OULU-NPU Protocol 1 | ACER: 6.3 -> 0.6 | Improved |
| add separate MT_v to evaluate MT_t (ablation) | MT-FAS · OULU-NPU Protocol 1 | ACER: 7.1 -> 0.6 | Improved |
| adapt surrogate detector to changing MT_t (ablation) | MT-FAS · OULU-NPU Protocol 1 | ACER: 4.2 -> 0.6 | Improved |
| momentum gamma 0.999 versus 0.9 (ablation) | MT_v update · OULU-NPU Protocol 1 | ACER: 5.3 -> 0.6 | Improved |
| swap meta-teacher backbone FAS-DR-Light for ResNet-8 (ablation) | MT-FAS · OULU-NPU Protocol 1 | ACER: 0.6 -> 0.8 | No clear effect |
Closest papers by meaning
- FaceShield: Explainable Face Anti-Spoofing with Multimodal Large Language Models (AAAI 2026 · similarity 0.88)
- SLIP: Spoof-Aware One-Class Face Anti-Spoofing with Language Image Pretraining (AAAI 2025 · similarity 0.86)
- From Intuition to Investigation: A Tool-Augmented Reasoning MLLM Framework for Generalizable Face Anti-Spoofing (CVPR 2026 · similarity 0.85)
- Interpretable Face Anti-Spoofing: Enhancing Generalization with Multimodal Large Language Models (AAAI 2025 · similarity 0.85)
- One-Class Face Anti-spoofing via Spoof Cue Map-Guided Feature Learning (CVPR 2024 · similarity 0.85)
- mmFAS: Multimodal Face Anti-Spoofing Using Multi-Level Alignment and Switch-Attention Fusion (AAAI 2025 · similarity 0.85)
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.
