Organ-level semantic segmentation on Wheat
85.54Spike IoUSPROUT
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| SPROUTPre-training Dataset=MCD-2.6M, Model Architecture=UDiT-L2026.03 | 85.54 | 51.58 | 86.44 | |
| SPROUTPre-training Dataset=MCD-2.6M, Model Architecture=UDiT-B2026.03 | 84.37 | 44.31 | 85 | |
| SPROUTPre-training Dataset=MCD-2.6M, Model Architecture=UDiT-S2026.03 | 82.43 | 38.23 | 83.52 | |
| DINOv2Pre-training Dataset=LVD-142M, Model Architecture=ViT-L-162026.03 | 81.77 | 37.05 | 81.46 | |
| CLIPPre-training Dataset=WIT-4M, Model Architecture=ViT-L-142026.03 | 81.3 | 32.43 | 81.51 | |
| DINOv2Pre-training Dataset=LVD-142M, Model Architecture=ViT-B-162026.03 | 81.05 | 34.8 | 80.96 | |
| DINOv3Pre-training Dataset=LVD-1689M, Model Architecture=ViT-L-162026.03 | 81.02 | 36.17 | 80.78 | |
| DINOv2Pre-training Dataset=LVD-142M, Model Architecture=ViT-S-162026.03 | 80.25 | 32.22 | 80.71 | |
| DINOv3Pre-training Dataset=LVD-1689M, Model Architecture=ViT-B-162026.03 | 79.44 | 31.21 | 79.98 | |
| CLIPPre-training Dataset=WIT-4M, Model Architecture=ViT-B-162026.03 | 78.3 | 26.32 | 78.88 | |
| MSNPre-training Dataset=ImageNet-1K, Model Architecture=ViT-S-162026.03 | 77.7 | 25.06 | 78.44 | |
| SigLIPPre-training Dataset=WebLI-10B, Model Architecture=ViT-L-162026.03 | 77.49 | 26.36 | 79.02 | |
| MSNPre-training Dataset=ImageNet-1K, Model Architecture=ViT-B-162026.03 | 76.81 | 24.01 | 78.16 | |
| SigLIPPre-training Dataset=WebLI-10B, Model Architecture=ViT-B-162026.03 | 76.77 | 25.2 | 78.52 | |
| DINOv3Pre-training Dataset=LVD-1689M, Model Architecture=ViT-S-162026.03 | 76.47 | 25.17 | 77.09 | |
| MSNPre-training Dataset=ImageNet-1K, Model Architecture=ViT-L-162026.03 | 76.17 | 20.63 | 77 | |
| MAEPre-training Dataset=ImageNet-1K, Model Architecture=ViT-L-162026.03 | 22.06 | 4.99 | 59.23 | |
| MAEPre-training Dataset=ImageNet-1K, Model Architecture=ViT-B-162026.03 | 21.28 | 2.08 | 59.86 |