Image-to-Text Retrieval on Wukong (test)
72.8R@1CLIP_ViT-L
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| CLIP_ViT-LBackbone=ViT-L, Evaluation Protocol=Fine-tuned2022.02 | 72.8 | 98.2 | 99.8 | 89.6 | |
| Wukong_ViT-LBackbone=ViT-L, Evaluation Protocol=Fine-tuned2022.02 | 68 | 94.4 | 98 | 85.8 | |
| CLIP_ViT-BBackbone=ViT-B, Evaluation Protocol=Fine-tuned2022.02 | 58.3 | 88.2 | 94.1 | 78.6 | |
| CLIP_Swin-LBackbone=Swin-L, Evaluation Protocol=Fine-tuned2022.02 | 56 | 86.1 | 92.5 | 76.7 | |
| Wukong_Swin-LBackbone=Swin-L, Evaluation Protocol=Fine-tuned2022.02 | 53.1 | 85.4 | 92.2 | 75 | |
| FILIP_ViT-LBackbone=ViT-L, Evaluation Protocol=Fine-tuned2022.02 | 53 | 85.3 | 92.7 | 76.3 | |
| FILIP_Swin-LBackbone=Swin-L, Evaluation Protocol=Fine-tuned2022.02 | 51 | 81.6 | 88.9 | 71.9 | |
| Wukong_ViT-BBackbone=ViT-B, Evaluation Protocol=Fine-tuned2022.02 | 50.5 | 82.7 | 90.5 | 73.3 | |
| FILIP_ViT-BBackbone=ViT-B, Evaluation Protocol=Fine-tuned2022.02 | 30.3 | 57.6 | 66.9 | 47.1 |