Image-Text Retrieval on RSICD (test)
10.43mRVSE++
Evaluation Results
| Method | Links | |
|---|---|---|
| VSE++Backbone=ResNet18, GRU, Params=15.78M2023.08 | 10.43 | |
| CAMP-bceBackbone=Faster R-CNN, biGRU, Params=27.03M2023.08 | 11.38 | |
| SCAN t-iBackbone=Faster R-CNN, biGRU, Params=13.68M2023.08 | 13.25 | |
| SCAN i2tParms=60 M2023.10 | 14.19 | |
| SCAN i-tBackbone=Faster R-CNN, biGRU, Params=13.68M2023.08 | 14.23 | |
| CAMP-tripletBackbone=Faster R-CNN, biGRU, Params=27.03M2023.08 | 14.39 | |
| CAMERAParms=64 M2023.10 | 14.39 | |
| LW-MCR-bBackbone=SqueezeNet, -, Params=1.65M2023.08 | 14.52 | |
| LW-MCR-dBackbone=SqueezeNet, -, Params=1.65M2023.08 | 14.66 | |
| LW-MCR2023.10 | 14.66 | |
| MTFNBackbone=Faster R-CNN, biGRU, Params=77.90M2023.08 | 14.81 | |
| VSE++Parms=29 M2023.10 | 14.89 | |
| Zero-shot CLIPBackbone=CLIP(ViT-B-32), Params=0.00M2023.08 | 15.04 | |
| CAMPParms=63 M2023.10 | 15.2 | |
| AMFMN-simBackbone=ResNet18, biGRU, Params=35.94M2023.08 | 15.53 | |
| AMFMNParms=36 M2023.10 | 15.53 | |
| LW-MCR-uBackbone=SqueezeNet, -, Params=1.65M2023.08 | 15.59 | |
| SCAN t2iParms=60 M2023.10 | 15.61 | |
| AMFMN-softBackbone=ResNet18, biGRU, Params=35.94M2023.08 | 16.02 | |
| AMFMN-fusionBackbone=ResNet18, biGRU, Params=35.94M2023.08 | 16.42 | |
| CoOpBackbone=CLIP(ViT-B-32), Params=0.10M2023.08 | 17.29 | |
| CABIRBackbone=ResNet152, BERT+biGRU, Params=Not specified2023.08 | 18.12 | |
| GaLR w/o MRBackbone=ResNet18, biGRU, Params=46.89M2023.08 | 18.62 | |
| MCRNBackbone=ResNet18, biGRU, Params=52.35M2023.08 | 18.95 | |
| GaLR with MRBackbone=ResNet18, biGRU, Params=46.89M2023.08 | 18.96 | |
| GaLRParms=46 M2023.10 | 18.96 | |
| VL-PromptBackbone=CLIP(ViT-B-32), Params=0.47M2023.08 | 19.61 | |
| KCR2023.10 | 19.89 | |
| HVSA2023.10 | 20.16 | |
| DOVE-SParms=25 M2023.10 | 20.4 | |
| SWANParms=40 M2023.10 | 20.61 | |
| ConvpassBackbone=CLIP(ViT-B-32), Params=0.14M2023.08 | 21.38 | |
| CLIP-AdapterBackbone=CLIP(ViT-B-32), Params=0.52M2023.08 | 21.65 | |
| VPTBackbone=CLIP(ViT-B-32), Params=0.46M2023.08 | 21.72 | |
| DOVEParms=38 M2023.10 | 22.72 | |
| Linear probeBackbone=CLIP(ViT-B-32), Params=1.05M2023.08 | 24.46 | |
| AdapterBackbone=CLIP(ViT-B-32), Params=0.17M2023.08 | 24.84 | |
| Single LanguageBackbone=CLIP(ViT-B-32), Params=151M2023.08 | 27.42 | |
| Cross-Modal AdapterBackbone=CLIP(ViT-B-32), Params=0.16M2023.08 | 27.96 | |
| AdaptFormerBackbone=CLIP(ViT-B-32), Params=0.17M2023.08 | 28.1 | |
| UniAdapterBackbone=CLIP(ViT-B-32), Params=0.55M2023.08 | 28.84 | |
| Full fine-tuningBackbone=CLIP(ViT-B-32), Params=151M2023.08 | 30.39 | |
| OursBackbone=CLIP(ViT-B-32), Params=0.16M2023.08 | 31.12 |