Research map / Vision-language and multimodal / Cross-modal learning
A Dataset and Benchmark for Large-Scale Multi-Modal Face Anti-Spoofing
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Replace halfway feature concatenation with modal-dependent squeeze-and-excitation feature re-weighting (ablation) | ResNet-18 multi-modal fusion · CASIA-SURF testing | TPR @FPR=10^-3 (%): 33.6 -> 81.8 | Helped |
| Replace halfway feature concatenation with modal-dependent squeeze-and-excitation feature re-weighting (ablation) | ResNet-18 multi-modal fusion · CASIA-SURF testing | ACER (%): 4.7 -> 2.4 | Helped |
| Use all three modalities (RGB, Depth, and IR) instead of any single modality (ablation) | ResNet-18 anti-spoofing model · CASIA-SURF testing | TPR @FPR=10^-4 (%): 6.8 -> 56.8 (RGB); 14.1 -> 56.8 (Depth); 10.9 -> 56.8 (IR) | Helped |
| Use all three modalities instead of any single modality (ablation) | ResNet-18 anti-spoofing model · CASIA-SURF testing | ACER (%): 11.3 -> 2.4 (RGB); 5.0 -> 2.4 (Depth); 8.1 -> 2.4 (IR) | Helped |
| Use two modalities instead of all three (ablation) | ResNet-18 anti-spoofing model · CASIA-SURF testing | TPR @FPR=10^-4 (%): 10.6 (RGB&Depth); 26.1 (RGB&IR); 24.3 (Depth&IR) -> 56.8 | Hurt |
| Pre-train FAS-TD-SF on CASIA-SURF before fine-tuning on SiW | FAS-TD-SF · SiW Protocol 1 | ACER (%): 1.05 -> 0.80 | Helped |
| Pre-train FAS-TD-SF on CASIA-SURF before fine-tuning on SiW | FAS-TD-SF · SiW Protocol 2 | ACER (%): 0.31$\pm$0.28 -> 0.17$\pm$0.16 | Helped |
| Pre-train FAS-TD-SF on CASIA-SURF before fine-tuning on SiW | FAS-TD-SF · SiW Protocol 3 | ACER (%): 7.73$\pm$3.99 -> 6.35$\pm$4.39 | Helped |
| Train FAS-TD-SF on CASIA-SURF rather than SiW | FAS-TD-SF · CASIA-MFSD | HTER (%): 39.4 -> 37.3 | Helped |
| Increase the number of training subjects (ablation) | CASIA-SURF anti-spoofing baseline · CASIA-SURF testing | TPR | Helped |
Closest papers by meaning
- FaceShield: Explainable Face Anti-Spoofing with Multimodal Large Language Models (AAAI 2026 · similarity 0.90)
- mmFAS: Multimodal Face Anti-Spoofing Using Multi-Level Alignment and Switch-Attention Fusion (AAAI 2025 · similarity 0.89)
- Suppress and Rebalance: Towards Generalized Multi-Modal Face Anti-Spoofing (CVPR 2024 · similarity 0.88)
- One-Class Face Anti-spoofing via Spoof Cue Map-Guided Feature Learning (CVPR 2024 · similarity 0.88)
- Interpretable Face Anti-Spoofing: Enhancing Generalization with Multimodal Large Language Models (AAAI 2025 · similarity 0.87)
- From Intuition to Investigation: A Tool-Augmented Reasoning MLLM Framework for Generalizable Face Anti-Spoofing (CVPR 2026 · 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.