Model Attribution on GM-CHQ (test)
64.8AccuracyManiFPT_FREQ
Evaluation results
| Method | Links | ||
|---|---|---|---|
| ManiFPT_FREQEmbedding Space=FREQ, Embedding map=Channelwise FFT2024.02 | 64.8 | 70.1 | |
| ManiFPT_RGBEmbedding Space=RGB, Embedding map=Identity2024.02 | 63.7 | 64.2 | |
| ManiFPT_SLEmbedding Space=SL, Embedding map=Pretrained ResNet502024.02 | 62.3 | 77.2 | |
| ManiFPT_SSLEmbedding Space=SSL, Embedding map=Pretrained BarlowTwin2024.02 | 61.9 | 63.3 | |
| Yu19Type=Learned representation2024.02 | 61.1 | 74.5 | |
| Wang20Type=Learned representation2024.02 | 59.5 | 30.3 | |
| Durall20Type=Frequency-based2024.02 | 59.1 | 38.8 | |
| Marra19Type=Learned representation2024.02 | 59.1 | 34.9 | |
| McClo18Type=Color-based2024.02 | 57.4 | 36.3 | |
| Dzanic20Type=Frequency-based2024.02 | 56.9 | 38.2 | |
| Nataraj19Type=Color-based2024.02 | 56.3 | 37.9 | |
| Marra18Type=Learned representation2024.02 | 51.3 | 20.5 |