Text-to-Image Retrieval on XTD (test)
95.4Recall@10 (EN)AltCLIP M9
Evaluation results
| Method | Links | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AltCLIP M9Backbone=ViT-L/14, Image encoder source=OpenAI2022.11 | 95.4 | — | — | — | — | — | — | — | 94.1 | 92.9 | 95.1 | 94.2 | 94.4 | 91.8 | 91.7 | |
| M-CLIP ViT-B/16+Backbone=ViT-B/16, Image encoder source=OpenCLIP2022.11 | 95 | — | — | — | — | — | — | — | 93.6 | 93.1 | 94 | 93.1 | 89 | 90 | 84.2 | |
| CLIP ViT-B/16+Backbone=ViT-B/16, Image encoder source=OpenCLIP2022.11 | 94.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| M-CLIP ViT-L/14Backbone=ViT-L/14, Image encoder source=OpenAI2022.11 | 92.4 | — | — | — | — | — | — | — | 91 | 90 | 89.7 | 91.1 | 85.2 | 85.8 | 81.9 | |
| CLIP ViT-L/14Backbone=ViT-L/14, Image encoder source=OpenAI2022.11 | 91.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| M-CLIP ViT-B/32Backbone=ViT-B/32, Image encoder source=OpenAI2022.11 | 91.8 | — | — | — | — | — | — | — | 89.1 | 89.4 | 89.3 | 89.8 | 82.1 | 86.1 | 81 | |
| CLIP ViT-B/32Backbone=ViT-B/32, Image encoder source=OpenAI2022.11 | 90.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| mUSE m32022.11 | 85.3 | — | — | — | — | — | — | — | 78.9 | 78.9 | 76.7 | 73.6 | 67.8 | 76.1 | 70.7 | |
| mUSE PATR2022.11 | 83.6 | — | — | — | — | — | — | — | 75.6 | 76.9 | 76.1 | 73.4 | 64.3 | 73.6 | 69.4 | |
| ALIGN-BASEmode=zero-shot, metric=Recall@102021.09 | — | 87.9 | 88.8 | 82.3 | 86.5 | 79.8 | 73.5 | 76.6 | — | — | — | — | — | — | — | |
| MURAL-BASEmode=zero-shot, metric=Recall@102021.09 | — | 88.4 | 89.6 | 83.6 | 88.3 | 86.1 | 84.8 | 82.4 | — | — | — | — | — | — | — | |
| MURAL-LARGEmode=zero-shot, metric=Recall@102021.09 | — | 91.8 | 92.9 | 87.2 | 89.7 | 91 | 89.5 | 88.1 | — | — | — | — | — | — | — | |
| mUSE+M3Lmode=zero-shot, metric=Recall@102021.09 | — | 78.9 | 76.7 | 73.6 | 76.1 | 71.7 | 70.9 | 70.7 | — | — | — | — | — | — | — |