Image-to-Text Retrieval on MS-COCO 1K
78.3R@1REPAIR
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| REPAIRNoise=20%2024.03 | 78.3 | 96.8 | 98.3 | |
| MSCNNoise=20%2024.03 | 78.1 | 97.2 | 98.8 | |
| NCRNoise=20%, re-implemented=true2024.03 | 77.7 | 95.8 | 98.3 | |
| REPAIRNoise=40%2024.03 | 76.4 | 95.4 | 98.4 | |
| NCRNoise=40%, re-implemented=true2024.03 | 76 | 95.1 | 98.4 | |
| DECLNoise=20%2024.03 | 75.6 | 95.1 | 98.3 | |
| BiCroNoise=20%, re-implemented=true2024.03 | 75.5 | 95.2 | 98.1 | |
| BiCroNoise=40%, re-implemented=true2024.03 | 75.5 | 95.2 | 98 | |
| MSCNNoise=40%2024.03 | 74.8 | 94.9 | 98 | |
| BiCroNoise=60%, re-implemented=true2024.03 | 74.7 | 94.2 | 97.9 | |
| MSCNNoise=60%2024.03 | 74.4 | 95.1 | 97.9 | |
| REPAIRNoise=60%2024.03 | 73.7 | 94.4 | 97.9 | |
| DECLNoise=40%2024.03 | 73.6 | 94.6 | 97.9 | |
| VSE∞Noise=20%2024.03 | 73.5 | 93.3 | 97 | |
| NCRNoise=60%, re-implemented=true2024.03 | 72.8 | 94 | 97.4 | |
| DECLNoise=60%2024.03 | 69.7 | 93.4 | 97.5 | |
| SEPSFine-grained alignment=true, Zero-shot=true, Backbone=CLIP ViT-B/16, Training Dataset=Flickr2025.11 | 65.8 | 77.4 | — | |
| VSE∞Noise=40%2024.03 | 53.3 | 84.3 | 92.1 | |
| SGRNoise=20%2024.03 | 49.1 | 83.8 | 92.7 | |
| LAPSFine-grained alignment=true, Zero-shot=true, Backbone=CLIP ViT-B/16, Training Dataset=Flickr2025.11 | 47.4 | 74.9 | — | |
| SGRNoise=40%2024.03 | 43.9 | 78.3 | 89.3 | |
| SGRNoise=60%2024.03 | 37.6 | 73.3 | 86.3 | |
| VSE∞Noise=60%2024.03 | 33.4 | 64.8 | 79.1 | |
| D2S-VSEFine-grained alignment=false, Zero-shot=true, Backbone=CLIP ViT-B/16, Training Dataset=Flickr2025.11 | 32.3 | 62.3 | — | |
| SCANFine-grained alignment=true, Zero-shot=true, Backbone=CLIP ViT-B/16, Training Dataset=Flickr2025.11 | 30.5 | 59.8 | — | |
| SCANNoise=40%2024.03 | 30.1 | 65.2 | 79.2 | |
| SCANNoise=20%2024.03 | 28.9 | 64.5 | 79.5 | |
| VSE++Fine-grained alignment=false, Zero-shot=true, Backbone=CLIP ViT-B/16, Training Dataset=Flickr2025.11 | 28.1 | 56.4 | — | |
| SCANNoise=60%2024.03 | 27.8 | 59.8 | 74.8 |