Scene Classification on EuroSAT Sentinel-2 (train val)
99.3OAIMP
Evaluation results
| Method | Links | |
|---|---|---|
| IMPBackbone=InternImage-XL, Pretraining=ImageNet-22K2024.03 | 99.3 | |
| IMP + MTPBackbone=InternImage-XL, Pretraining=ImageNet-22K2024.03 | 99.24 | |
| SpectralGPT+Backbone=ViT-B2024.03 | 99.21 | |
| SatMAE++Backbone=ViT-L2024.03 | 99.04 | |
| MAE + MTPBackbone=ViT-L + RVSA2024.03 | 98.78 | |
| MAE + MTPBackbone=ViT-B + RVSA2024.03 | 98.76 | |
| CtxMIMBackbone=Swin-B2024.03 | 98.69 | |
| Scale-MAEBackbone=ViT-L2024.03 | 98.59 | |
| MAEBackbone=ViT-L + RVSA2024.03 | 98.56 | |
| MAEBackbone=ViT-B + RVSA2024.03 | 98.54 | |
| UsatBackbone=ViT-L2024.03 | 98.37 | |
| CACoBackbone=ResNet-502024.03 | 97.77 | |
| SeCoBackbone=ResNet-502024.03 | 97.34 | |
| GASSLBackbone=ResNet-502024.03 | 96.38 | |
| SwiMDiffBackbone=ResNet-182024.03 | 96.1 | |
| SatMAEBackbone=ViT-L2024.03 | 95.74 | |
| SeCoBackbone=ResNet-182024.03 | 93.14 | |
| GASSLBackbone=ResNet-182024.03 | 89.51 |