Image-to-Text Retrieval on CC3M
47.2RecallCLIP-RD
Evaluation Results
| Method | Links | |
|---|---|---|
| CLIP-RDBackbone=ResNet-50, Zero-shot=true2026.03 | 47.2 | |
| CLIP-KDBackbone=ResNet-50, Zero-shot=true2026.03 | 47.1 | |
| ViT-B/16Backbone=ViT-B/16, Zero-shot=true, Role=Teacher2026.03 | 43.8 | |
| CLIP-RDBackbone=EfficientNet-B0, Zero-shot=true2026.03 | 42.8 | |
| ResNet-50Backbone=ResNet-50, Zero-shot=true, Role=Student2026.03 | 40.7 | |
| CLIP-RDBackbone=ResNet-18, Zero-shot=true2026.03 | 37.6 | |
| CLIP-KDBackbone=ResNet-18, Zero-shot=true2026.03 | 36.9 | |
| EfficientNet-B0Backbone=EfficientNet-B0, Zero-shot=true, Role=Student2026.03 | 36.3 | |
| ResNet-18Backbone=ResNet-18, Zero-shot=true, Role=Student2026.03 | 31.7 |