Text-to-Image Retrieval on ICR
61.6R@1Chinese CLIP
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Chinese CLIPBackbone=ViT-L/14, Evaluation Protocol=Finetuned2022.11 | 61.6 | 83.6 | 89 | |
| R2D2Backbone=ViT-L/14, Evaluation Protocol=Finetuned2022.11 | 60.7 | 82 | 86.9 | |
| Chinese CLIPBackbone=ViT-B/16, Evaluation Protocol=Finetuned2022.11 | 55.4 | 79 | 85.2 | |
| R2D2Backbone=ViT-B, Evaluation Protocol=Finetuned2022.11 | 42.2 | 69.4 | 77.8 |