Morphing Attack Detection on MD #3
4D-EERProposed Method
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Proposed MethodTrain:Val Protocol=MD#2:MD#12026.04 | 4 | 16.59 | 4.61 | 3.23 | |
| Proposed MethodTraining/Validation Protocol=Train:Val MD#1:MD#2, Morph Types (MTs) Aggregation=Averaged over MTs2026.04 | 7.14 | 18.21 | 7.7 | 3.78 | |
| LocalMorphTraining/Validation Protocol=Train:Val MD#1:MD#2, Morph Types (MTs) Aggregation=Averaged over MTs2026.04 | 12.74 | 22.13 | 10.78 | 6.86 | |
| Distil-MADTrain:Val Protocol=MD#2:MD#12026.04 | 13.72 | 66.67 | 26.47 | 15.69 | |
| ViTSVMTraining/Validation Protocol=Train:Val MD#1:MD#2, Morph Types (MTs) Aggregation=Averaged over MTs2026.04 | 14.95 | 44.4 | 28.43 | 21.29 | |
| Distil-MADTraining/Validation Protocol=Train:Val MD#1:MD#2, Morph Types (MTs) Aggregation=Averaged over MTs2026.04 | 16.57 | 60.78 | 38.24 | 28.43 | |
| LocalMorphTrain:Val Protocol=MD#2:MD#12026.04 | 27.45 | 68.13 | 59.12 | 46.27 | |
| Self-SupervisionTrain:Val Protocol=MD#2:MD#12026.04 | 28.93 | 73.81 | 57.98 | 49.86 | |
| FS-SPNTraining/Validation Protocol=Train:Val MD#1:MD#2, Morph Types (MTs) Aggregation=Averaged over MTs2026.04 | 29.41 | 89.21 | 72.54 | 51.96 | |
| Self-SupervisionTraining/Validation Protocol=Train:Val MD#1:MD#2, Morph Types (MTs) Aggregation=Averaged over MTs2026.04 | 30.89 | 65.27 | 51.26 | 46.78 | |
| MADationTrain:Val Protocol=MD#2:MD#12026.04 | 31.03 | 80.25 | 69.05 | 57.42 | |
| FS-SPNTrain:Val Protocol=MD#2:MD#12026.04 | 31.51 | 94.11 | 82.35 | 60.78 | |
| Few-Shot MADBackbone=MobileNet, Training/Validation Protocol=Train:Val MD#1:MD#2, Morph Types (MTs) Aggregation=Averaged over MTs2026.04 | 33.6 | 80.39 | 69.89 | 59.38 | |
| Few-Shot MAD (MobileNet)Backbone=MobileNet, Train:Val Protocol=MD#2:MD#12026.04 | 33.6 | 80.39 | 69.89 | 59.38 | |
| ViTSVMTrain:Val Protocol=MD#2:MD#12026.04 | 42.73 | 90.48 | 78.43 | 71.85 | |
| Few-Shot MADBackbone=ResNet, Training/Validation Protocol=Train:Val MD#1:MD#2, Morph Types (MTs) Aggregation=Averaged over MTs2026.04 | 48.85 | 94.96 | 92.3 | 86.97 | |
| Few-Shot MAD (ResNet)Backbone=ResNet, Train:Val Protocol=MD#2:MD#12026.04 | 48.85 | 94.96 | 92.3 | 86.97 | |
| MADationTraining/Validation Protocol=Train:Val MD#1:MD#2, Morph Types (MTs) Aggregation=Averaged over MTs2026.04 | 49.6 | 65.27 | 56.3 | 47.34 | |
| Diffusion probabilistic modelsTraining/Validation Protocol=Train:Val MD#1:MD#2, Morph Types (MTs) Aggregation=Averaged over MTs2026.04 | 57.07 | 100 | 99.58 | 98.04 | |
| Diffusion probabilistic modelsTrain:Val Protocol=MD#2:MD#12026.04 | 66.67 | 100 | 100 | 100 |