Pixel-wise classification on Burn Scars PANGAEA benchmark
47mIoUMicro UNet
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Micro UNetTunable Parameters (M)=0.492026.04 | 47 | 62 | 62 | |
| LIANet-ModifiedTunable Parameters (M)=0.872026.04 | 47 | 63 | 63 | |
| UNetTunable Parameters (M)=17.32026.04 | 45 | 65 | 62 | |
| Prithvi v2-300Setting=Full, Tunable Parameters (M)=3132026.04 | 44 | 60 | 59 | |
| Prithvi v2-300Setting=Frozen, Tunable Parameters (M)=7.72026.04 | 43 | 63 | 59 | |
| Prithvi v2-300Setting=Embedding, Tunable Parameters (M)=0.462026.04 | 42 | 61 | 58 | |
| TerraMind-baseSetting=Frozen, Tunable Parameters (M)=5.02026.04 | 41 | 60 | 57 | |
| DOFA-LargeSetting=Frozen, Tunable Parameters (M)=7.72026.04 | 41 | 59 | 58 | |
| TerraMind-baseSetting=Full, Tunable Parameters (M)=922026.04 | 39 | 58 | 55 | |
| DOFA-LargeSetting=Embedding, Tunable Parameters (M)=0.462026.04 | 38 | 55 | 55 | |
| DOFA-LargeSetting=Full, Tunable Parameters (M)=3572026.04 | 37 | 59 | 51 | |
| TerraMind-baseSetting=Embedding, Tunable Parameters (M)=0.392026.04 | 35 | 52 | 51 |