Image-to-Text Retrieval on Flickr8K CN
83.3R@1Wukong_ViT-L
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Wukong_ViT-LBackbone=ViT-L, Evaluation Protocol=Fine-tuned2022.02 | 83.3 | 97.3 | 99.5 | 89.7 | |
| CLIP_ViT-LBackbone=ViT-L, Evaluation Protocol=Fine-tuned2022.02 | 81.4 | 96.9 | 99 | 88.6 | |
| FILIP_ViT-LBackbone=ViT-L, Evaluation Protocol=Fine-tuned2022.02 | 80.8 | 94.8 | 98.3 | 88 | |
| CLIP_ViT-BBackbone=ViT-B, Evaluation Protocol=Fine-tuned2022.02 | 77.7 | 94.7 | 98.1 | 85.3 | |
| FILIP_Swin-LBackbone=Swin-L, Evaluation Protocol=Fine-tuned2022.02 | 77.6 | 94.4 | 97.7 | 85.1 | |
| CLIP_Swin-LBackbone=Swin-L, Evaluation Protocol=Fine-tuned2022.02 | 77.3 | 94.9 | 98.2 | 84.8 | |
| Wukong_Swin-LBackbone=Swin-L, Evaluation Protocol=Fine-tuned2022.02 | 74.9 | 93.6 | 97.8 | 83.6 | |
| Wukong_ViT-BBackbone=ViT-B, Evaluation Protocol=Fine-tuned2022.02 | 71.7 | 91.5 | 96.6 | 82.6 | |
| FILIPBackbone=ViT-L, Zero-shot=true2022.02 | 70 | 91.6 | 96.6 | 79.8 | |
| CLIPBackbone=ViT-L, Zero-shot=true2022.02 | 65.4 | 89.2 | 95.4 | 77.2 | |
| WukongBackbone=ViT-L, Zero-shot=true2022.02 | 61.4 | 86.2 | 93.6 | 74.4 | |
| CLIPBackbone=ViT-B, Zero-shot=true2022.02 | 59.5 | 86.2 | 93.4 | 72.7 | |
| CLIPBackbone=Swin-L, Zero-shot=true2022.02 | 56 | 83.2 | 92.4 | 69.2 | |
| WukongBackbone=ViT-B, Zero-shot=true2022.02 | 55.4 | 82.3 | 90 | 70.6 | |
| FILIP_ViT-BBackbone=ViT-B, Evaluation Protocol=Fine-tuned2022.02 | 52.6 | 81.5 | 90.2 | 72.4 | |
| FILIPBackbone=Swin-L, Zero-shot=true2022.02 | 52.4 | 78 | 87.2 | 67.7 | |
| WukongBackbone=Swin-L, Zero-shot=true2022.02 | 47.2 | 78.8 | 87.6 | 65.2 | |
| FILIPBackbone=ViT-B, Zero-shot=true2022.02 | 37.2 | 65.9 | 75.2 | 52.5 | |
| BriVLBackbone=N/A, Zero-shot=true2022.02 | 13.4 | 31.2 | 40.7 | 23.9 |