Appendix D: Research Papers & Recommended Reading
Face Anti-Spoofing (Core)
- FaceForensics++ (Rössler et al., ICCV 2019) — Benchmark dataset for face manipulation detection
- CDCN (Yu et al., CVPR 2020) — Central Difference Convolutional Network for face anti-spoofing
- NAS-FAS (Yu et al., TPAMI 2020) — Neural architecture search for face anti-spoofing
- SSDG (Jia et al., CVPR 2020) — Single Side Domain Generalization
- SSAN (Wang et al., CVPR 2022) — Shuffled Style Assembly Network
Domain Generalization
- MADDG (Shao et al., CVPR 2019) — Multi-Adversarial Discriminative Deep Domain Generalization
- DRDG (Liu et al., AAAI 2021) — Dual Reweighting Domain Generalization
- AMEL (Zhou et al., AAAI 2022) — Adaptive Meta-learning
Deepfake Detection
- Thinking in Frequency (Qian et al., ECCV 2020)
- Multi-Attentional Deepfake Detection (Zhao et al., CVPR 2021)
- DeepfakeBench (Yan et al., NeurIPS 2023)
- Implicit Identity Leakage (Dong et al., CVPR 2023)
Datasets
- OULU-NPU (Boulkenafet et al., 2017) — 4 protocols, 4 conditions
- CASIA-FASD (Zhang et al., 2012) — Classic benchmark
- SiW (Liu et al., CVPR 2018) — Spoof in the Wild
- CelebA-Spoof (Zhang et al., ECCV 2020) — Large-scale with rich annotations
- WMCA (George et al., TIFS 2020) — Wide Multi-Channel Attack database
- SiW-Mv2 (Guo et al., 2022) — Extended spoof in the wild