EN中文

Figure/Table detail

MDSNet : A Multi-Scale Depth Supervision Method for High-Resolution Remote Sensing Image Semantic Segmentation
SHAN Huilin, WANG Xingtao, LIU Wenxing, WU Xinyue, GAO Runze, LI Hongxu
Journal of Geo-information Science, 2025, 27(6): 1381-1400.   DOI: 10.12082/dqxxkx.2025.250009

模型 不透表面 建筑物 低矮植被 树木 汽车 mIoU mF1 OA
IoU F1 IoU F1 IoU F1 IoU F1 IoU F1
U-Net 78.89 88.20 88.47 93.88 71.81 83.59 71.82 83.60 79.92 88.84 71.81 82.53 86.70
EIGNet 80.07 88.93 89.02 94.19 71.28 83.23 69.25 81.83 78.27 87.81 73.29 84.05 86.91
DeepLabV3+ 81.93 90.07 89.76 94.60 72.59 84.12 68.98 81.64 78.19 87.76 74.67 85.07 87.72
MAResU-Net 81.42 89.76 90.01 94.74 72.46 84.03 70.10 82.42 78.43 87.71 74.70 85.08 87.76
CMLFormer 82.87 90.63 70.79 95.17 73.26 84.56 69.99 82.34 79.51 88.58 75.55 85.64 88.33
A2FPN 83.29 90.88 90.96 95.26 73.79 84.92 70.38 82.61 78.86 88.18 76.18 86.12 88.63
CGGLNet 83.85 91.21 91.68 95.66 74.64 85.48 71.57 83.43 80.93 89.46 76.98 86.60 89.09
CMTFNet 84.29 91.48 91.58 95.60 74.90 85.65 71.80 83.59 80.66 89.29 77.61 87.08 89.31
PSPNet 84.98 91.88 92.34 96.02 75.75 86.20 72.21 83.86 81.16 89.60 78.01 87.30 89.68
MDSNet 88.60 93.96 94.47 97.16 79.67 88.68 76.54 86.71 84.78 91.76 82.43 90.16 91.94
Tab. 5 Segmentation results under different network models on the Potsdam dataset (%)
Other figure/table from this article