Multi-label Image Classification on NUS-WIDE
73.35Micro F1Sim-Dissim
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Sim-DissimBackbone=ResNet-502024.10 | 73.35 | 57.49 | 56.74 | |
| MSCBackbone=ResNet-502024.10 | 72.02 | 56.13 | 52.87 | |
| MulSupConBackbone=ResNet-502024.10 | 71.88 | 54.36 | 52.47 | |
| Class protoBackbone=ResNet-502024.10 | 71.79 | 56.03 | 52.95 | |
| JaccardBackbone=ResNet-502024.10 | 71.07 | 52.84 | 51.25 | |
| ALLBackbone=ResNet-502024.10 | 70.25 | 52.84 | 51.35 | |
| ANYBackbone=ResNet-502024.10 | 68.42 | 50.65 | 49.28 |