Image-to-Text Retrieval on RSICD (test)
15.65R@1LRSCLIP
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| LRSCLIPBackbone=ViT-L/142026.03 | 15.65 | 31.38 | 41.26 | |
| GeoAlignCLIPBackbone=ViT-L/142026.03 | 14.82 | 30.38 | 40.71 | |
| GeoAlignCLIPBackbone=ViT-B/162026.03 | 14 | 27.81 | 38.06 | |
| LRSCLIPBackbone=ViT-B/162026.03 | 13.08 | 29.83 | 39.52 | |
| GeoRSCLIPBackbone=ViT-L/142026.03 | 12.81 | 28.91 | 41.17 | |
| GeoRSCLIPBackbone=ViT-L/14-3362026.03 | 12.08 | 27.26 | 39.71 | |
| FGCLIPBackbone=ViT-L/14-3362026.03 | 11.25 | 25.34 | 34.95 | |
| GeoRSCLIPBackbone=ViT-B/322026.03 | 11.25 | 26.53 | 37.15 | |
| FGCLIPBackbone=ViT-B/162026.03 | 9.42 | 21.41 | 30.56 | |
| LongCLIPBackbone=ViT-L/142026.03 | 8.78 | 20.4 | 28.82 | |
| DOVEParms=38 M2023.10 | 8.66 | 22.35 | 34.95 | |
| LongCLIPBackbone=ViT-B/162026.03 | 7.69 | 20.13 | 29.19 | |
| DOVE-SParms=25 M2023.10 | 7.62 | 19.85 | 30.65 | |
| HVSA2023.10 | 7.47 | 20.62 | 32.11 | |
| SWANParms=40 M2023.10 | 7.41 | 20.13 | 30.86 | |
| GaLRParms=46 M2023.10 | 6.59 | 19.85 | 31.04 | |
| CLIPBackbone=ViT-L/142026.03 | 6.59 | 17.75 | 28.09 | |
| SkyCLIPBackbone=ViT-B/322026.03 | 6.22 | 17.2 | 26.53 | |
| KCR2023.10 | 5.95 | 18.59 | 29.58 | |
| SkyCLIPBackbone=ViT-L/14-pct302026.03 | 5.86 | 18.39 | 27.08 | |
| SkyCLIPBackbone=ViT-L/14-pct502026.03 | 5.86 | 18.3 | 27.08 | |
| CLIPBackbone=ViT-B/162026.03 | 5.67 | 14.82 | 22.6 | |
| AMFMNParms=36 M2023.10 | 5.21 | 14.72 | 21.57 | |
| SCAN i2tParms=60 M2023.10 | 4.82 | 13.66 | 21.99 | |
| SCAN t2iParms=60 M2023.10 | 4.79 | 16.19 | 24.86 | |
| CAMPParms=63 M2023.10 | 4.64 | 14.61 | 24.09 | |
| CAMERAParms=64 M2023.10 | 4.57 | 13.08 | 21.77 | |
| VSE++Parms=29 M2023.10 | 4.56 | 16.73 | 22.94 | |
| LW-MCR2023.10 | 3.29 | 12.52 | 19.93 |