Blade Segmentation on Blade Segmentation Dataset (test)
97.61AccuracyBU-Netv2+P (ŝ)
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| BU-Netv2+P (ŝ)2026.03 | 97.61 | 97.67 | 96.5 | 94.58 | 1.13 | 1.42 | |
| BU-Netv1+P2026.03 | 97.39 | 93.35 | 95.73 | 93.8 | 1.13 | 1.35 | |
| Mask2Former2026.03 | 96.68 | 93.89 | 94.76 | 93.72 | 1.12 | 1.37 | |
| EfficientFormer2026.03 | 96.42 | 93.63 | 94.55 | 93.81 | 1.12 | 1.37 | |
| DiffSegRegion-only proposal=true2026.03 | 96.37 | 85.73 | 86.4 | 72.14 | 1.05 | 1.12 | |
| U-NetFormerBackbone architecture=ResNet-182026.03 | 96.2 | 93.51 | 94.42 | 91.75 | 1.12 | 1.36 | |
| MobileViT2026.03 | 96.14 | 93.33 | 94.38 | 93.47 | 1.11 | 1.36 | |
| BiRefNet2026.03 | 95.65 | 92.52 | 94.57 | 92.38 | 1.11 | 1.35 | |
| SAMRegion-only proposal=true2026.03 | 94.36 | 91.22 | 92.6 | 91.66 | 1.1 | 1.34 | |
| ResNeStBackbone architecture=ResNet-502026.03 | 94.23 | 91.47 | 92.77 | 89.63 | 1.09 | 1.33 | |
| DeepLabv3+Backbone architecture=ResNet-502026.03 | 94.14 | 87.38 | 89.03 | 87.47 | 1.09 | 1.27 | |
| SWBackbone architecture=ResNet-502026.03 | 93.48 | 91.71 | 91.37 | 87.44 | 1.08 | 1.33 | |
| U-Net2026.03 | 86.24 | 68.93 | 77.95 | 75.94 | 1 | 1 |