Text-to-Image Retrieval on MSCOCO CN (5K)
6,130R@1AltCLIP
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| AltCLIPBackbone=ViT-L, Zero-shot=true, Stage=Contrastive Learning2022.11 | 6,130 | 86 | 93.2 | 85 | |
| CN-CLIPBackbone=ViT-L, Zero-shot=true2022.11 | 5,860 | 85.3 | 92.7 | 82.4 | |
| AltCLIPTBackbone=ViT-L, Zero-shot=true, Stage=Teacher Learning2022.11 | 5,490 | 81.8 | 90.5 | 82.5 | |
| TaiyiBackbone=ViT-L, Zero-shot=true2022.11 | 4,610 | 74.9 | 85.1 | 73.3 | |
| CLIPBackbone=ViT-L, Zero-shot=true2022.11 | 80 | 3.9 | 5.8 | 6.2 |