Text-to-Image Retrieval on Flickr30k EN
72.5R@1AltCLIP
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| AltCLIPBackbone=ViT-L, Zero-shot=true, Stage=Contrastive Learning2022.11 | 72.5 | 91.6 | 95.4 | 90.4 | |
| AltCLIPTBackbone=ViT-L, Zero-shot=true, Stage=Teacher Learning2022.11 | 66.3 | 87.8 | 92.7 | 88.3 | |
| CLIPBackbone=ViT-L, Zero-shot=true2022.11 | 65 | 87.1 | 92.2 | 87.6 | |
| CN-CLIPBackbone=ViT-L, Zero-shot=true2022.11 | 49.5 | 76.9 | 83.8 | 77.3 | |
| TaiyiBackbone=ViT-L, Zero-shot=true2022.11 | 25.3 | 48.2 | 59.2 | 53.3 |