Image Classification on ImageNet (val) (CR and Accuracy Loss)
128Contrastive Ratio (CR)LegoNet-C
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LegoNet-CBackbone=ResNet-502026.02 | 128 | 2.8 | |
| LegoNet-ABackbone=ResNet-502026.02 | 64 | 0 | |
| DCBackbone=ResNet-502026.02 | 49 | 0 | |
| LSSQBackbone=ResNet-502026.02 | 49 | 3.2 | |
| ARQBackbone=ResNet-502026.02 | 26 | 0 | |
| VQBackbone=ResNet-502026.02 | 24 | 1 | |
| PQFTBackbone=ResNet-502026.02 | 19 | 1.57 | |
| DKM4Backbone=ResNet-502026.02 | 16 | 1.3 | |
| MMSEBackbone=ResNet-502026.02 | 8 | 1.7 | |
| ANBackbone=ResNet-502026.02 | 3.8 | 0.67 | |
| CHIPBackbone=ResNet-502026.02 | 2.26 | 0.89 | |
| DMCBackbone=ResNet-502026.02 | 2.22 | 0.8 | |
| CCBackbone=ResNet-502026.02 | 2.07 | 0.56 |