Table 2 Prediction results for nanoparticle data on each model.
From: The segmentation of nanoparticles with a novel approach of HRU2-Net†
Models | Backbone | GFOPs | Params | MIoU (%) | Acc (%) | Kappa (%) | Dice (%) |
|---|---|---|---|---|---|---|---|
HRU2-Net† | – | 42.60 | 6.57 | 87.37 | 97.31 | 85.95 | 92.98 |
U-HRNet | HRNet48 | – | – | 87.21 | 97.49 | 85.72 | 95.86 |
U2-Net† | – | 48.77 | 4.36 | 87.29 | 97.47 | 85.83 | 92.91 |
U-Net | – | 124.31 | 51.14 | 86.65 | 97.08 | 85.07 | 92.53 |
PSPNet | ResNet50 | 265.5 | 259.03 | 86.72 | 97.33 | 85.11 | 92.56 |
DDRNet | DdrNet_23 | 17.93 | 77.00 | 85.28 | 97.01 | 83.28 | 91.64 |
PPLiteSeg | STDC2 | 9.04 | 46.07 | 85.77 | 97.12 | 83.91 | 91.95 |
RTFormer | – | 16.90 | 64.36 | 85.77 | 97.01 | 83.92 | 91.96 |
Bisenetv2 | – | 8.06 | 8.88 | 83.96 | 96.48 | 81.61 | 90.80 |
SwinUnet | – | – | – | 84.07 | 96.40 | 96.40 | 92.22 |
Efficientformerv2 | – | – | – | 84.57 | 96.61 | 82.42 | 91.20 |
