Cross-modal Retrieval on RSITMD (test)
32.74R@1 (Image-to-Text)HarMA
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| HarMABackbone (image/text)=GeoRSCLIP(ViT-B-32-RET-2), Trainable Params=0.50M, Extra Data=false2024.04 | 32.74 | 53.76 | 69.25 | 25.62 | 57.65 | 74.6 | 52.27 | |
| Full-FT GeoRSCLIPBackbone (image/text)=GeoRSCLIP(ViT-B-32-RET-2), Trainable Params=151M, Extra Data=true, Evaluation Protocol=Full Fine-tuned2024.04 | 32.3 | 53.32 | 67.92 | 25.04 | 57.88 | 74.38 | 51.81 | |
| Full-FT GeoRSCLIPBackbone (image/text)=GeoRSCLIP(ViT-B-32-RET-2), Trainable Params=151M, Evaluation Protocol=Full Fine-tuned2024.04 | 30.53 | 49.78 | 63.05 | 24.91 | 57.21 | 75.35 | 50.14 | |
| Full-FT CLIPBackbone (image/text)=CLIP(ViT-B-32), Trainable Params=151M, Evaluation Protocol=Full Fine-tuned2024.04 | 26.99 | 46.9 | 58.85 | 20.53 | 52.35 | 71.15 | 46.13 | |
| HarMABackbone (image/text)=CLIP(ViT-B-32), Trainable Params=0.50M, Extra Data=false2024.04 | 25.81 | 48.37 | 60.61 | 19.92 | 53.27 | 71.21 | 46.53 | |
| PE-RSITRBackbone (image/text)=CLIP(ViT-B-32), Trainable Params=0.16M, Evaluation Protocol=Parameter-efficient Fine-tuned2024.04 | 23.67 | 44.07 | 60.36 | 20.1 | 50.63 | 67.97 | 44.47 | |
| UniAdapterBackbone (image/text)=CLIP(ViT-B-32), Trainable Params=0.55M, Evaluation Protocol=Parameter-efficient Fine-tuned2024.04 | 19.86 | 36.32 | 51.28 | 17.54 | 44.89 | 56.46 | 39.23 | |
| Cross-Modal AdapterBackbone (image/text)=CLIP(ViT-B-32), Trainable Params=0.16M, Evaluation Protocol=Parameter-efficient Fine-tuned2024.04 | 18.16 | 36.08 | 48.72 | 16.31 | 44.33 | 64.75 | 38.06 | |
| AdaptFormerBackbone (image/text)=CLIP(ViT-B-32), Trainable Params=0.17M, Evaluation Protocol=Parameter-efficient Fine-tuned2024.04 | 16.71 | 30.16 | 42.91 | 14.27 | 41.53 | 61.46 | 34.81 | |
| CLIP-AdapterBackbone (image/text)=CLIP(ViT-B-32), Trainable Params=0.52M, Evaluation Protocol=Parameter-efficient Fine-tuned2024.04 | 12.83 | 28.84 | 39.05 | 13.3 | 40.2 | 60.06 | 32.38 | |
| RemoteCLIPTraining Dataset=RET-3 + DET-10 + SEG-4, Training Samples=165,745, Date=Jun 2023, Image Backbone Name=ViT-L-14, Image Backbone Params=304, Text Backbone Name=Transformer, Text Backbone Params=124, Total Params=4282023.06 | 0.2876 | 0.5243 | 0.6394 | 0.2376 | 0.5951 | 0.7473 | 0.5052 | |
| RemoteCLIPTraining Dataset=RET-3 + DET-10 + SEG-4, Training Samples=165,745, Date=Jun 2023, Image Backbone Name=ViT-B-32, Image Backbone Params=87, Text Backbone Name=Transformer, Text Backbone Params=63, Total Params=1512023.06 | 0.2788 | 0.5066 | 0.6571 | 0.2217 | 0.5646 | 0.7341 | 0.4938 | |
| CLIP-CLTraining Dataset=RET-3, Training Samples=13,713, Image Backbone Name=ViT-B-32, Image Backbone Params=88, Text Backbone Name=Transformer, Text Backbone Params=64, Total Params=1512023.06 | 0.2478 | 0.5 | 0.6327 | 0.2261 | 0.5527 | 0.6987 | 0.4763 | |
| RemoteCLIPTraining Dataset=RET-3 + DET-10 + SEG-4, Training Samples=165,745, Date=Jun 2023, Image Backbone Name=ResNet-50, Image Backbone Params=38, Text Backbone Name=Transformer, Text Backbone Params=64, Total Params=1022023.06 | 0.2367 | 0.4757 | 0.646 | 0.1929 | 0.5155 | 0.7058 | 0.4621 | |
| Rahhal et al.Training Dataset=RSITMD, Training Samples=4,743, Date=Oct 2022, Image Backbone Name=ViT-B-32, Image Backbone Params=87, Text Backbone Name=Transformer, Text Backbone Params=63, Total Params=1512023.06 | 0.1969 | 0.4026 | 0.5442 | 0.1761 | 0.4973 | 0.6659 | 0.4138 | |
| CLIP-CLTraining Dataset=RET-3, Training Samples=13,713, Image Backbone Name=ResNet-50, Image Backbone Params=38, Text Backbone Name=Transformer, Text Backbone Params=64, Total Params=1022023.06 | 0.1925 | 0.3982 | 0.5133 | 0.1509 | 0.4146 | 0.5664 | 0.3727 | |
| PIRTraining Dataset=RSITMD, Training Samples=4,743, Date=Oct 2023, Image Backbone Name=SwinT+ ResNet-50, Text Backbone Name=Bert2023.06 | 0.1814 | 0.4115 | 0.5288 | 0.1217 | 0.4168 | 0.6341 | 0.3824 | |
| DOVETraining Dataset=RSITMD, Training Samples=4,743, Date=Oct 2023, Image Backbone Name=ResNet-50, Text Backbone Name=GRU2023.06 | 0.1681 | 0.368 | 0.4993 | 0.122 | 0.4993 | 0.665 | 0.387 | |
| GaLRTraining Dataset=RSITMD, Training Samples=4,743, Date=Apr 2022, Image Backbone Name=ResNet-18, Text Backbone Name=GRU2023.06 | 0.1482 | 0.3164 | 0.4248 | 0.1115 | 0.3668 | 0.5168 | 0.3141 | |
| HVSATraining Dataset=RSITMD, Training Samples=4,743, Date=Sept 2023, Image Backbone Name=ResNet-18, Text Backbone Name=GRU2023.06 | 0.132 | 0.3208 | 0.4558 | 0.1143 | 0.392 | 0.5745 | 0.3316 | |
| FBCLMTraining Dataset=RSITMD, Training Samples=4,743, Date=May 2022, Image Backbone Name=Resnet-18, Text Backbone Name=BERT2023.06 | 0.1284 | 0.3053 | 0.4589 | 0.1044 | 0.3701 | 0.5794 | 0.3244 | |
| HyperMatchTraining Dataset=RSITMD, Training Samples=4,743, Date=Dec 2022, Image Backbone Name=ResNet-18, Text Backbone Name=GRU2023.06 | 0.1173 | 0.281 | 0.3805 | 0.0916 | 0.3231 | 0.4664 | 0.2766 | |
| SCANTraining Dataset=RSITMD, Training Samples=4,743, Date=Mar 2018, Image Backbone Name=ResNet-101, Text Backbone Name=GRU2023.06 | 0.1106 | 0.2588 | 0.3938 | 0.0982 | 0.2938 | 0.4212 | 0.2628 | |
| CMFM-NetTraining Dataset=RSITMD, Training Samples=4,743, Date=Dec 2022, Image Backbone Name=ResNet-18, Text Backbone Name=GRU2023.06 | 0.1084 | 0.2876 | 0.4004 | 0.1 | 0.3283 | 0.4721 | 0.2828 | |
| AMFMNTraining Dataset=RSITMD, Training Samples=4,743, Date=Aug 2020, Image Backbone Name=ResNet-50, Text Backbone Name=GloVe fasText2023.06 | 0.1063 | 0.2478 | 0.4181 | 0.1151 | 0.3469 | 0.5487 | 0.2972 | |
| MTFNTraining Dataset=RSITMD, Training Samples=4,743, Date=Aug 2019, Image Backbone Name=ResNet, Text Backbone Name=GRU2023.06 | 0.104 | 0.2765 | 0.3628 | 0.0996 | 0.3137 | 0.4584 | 0.2692 | |
| VSE++Training Dataset=RSITMD, Training Samples=4,743, Date=Jul 2017, Image Backbone Name=VGG19, Text Backbone Name=GRU2023.06 | 0.1038 | 0.2765 | 0.396 | 0.0779 | 0.2487 | 0.3867 | 0.2483 | |
| LW-MRC-uTraining Dataset=RSITMD, Training Samples=4,743, Date=Dec 2020, Image Backbone Name=Big Transfer, Text Backbone Name=Bi-LSTM2023.06 | 0.0973 | 0.2677 | 0.3761 | 0.0925 | 0.3407 | 0.5403 | 0.2858 |