Table 4 Quantitative comparisons with classical models on the INRIA dataset.

From: Dynamic atrous attention and dual branch context fusion for cross scale Building segmentation in high resolution remote sensing imagery

 

OA (%)

Precision (%)

Recall (%)

F1-score (%)

mIoU (%)

DeepLabV3+

97.84

95.46

93.42

94.43

89.74

PSPNet

97.42

94.90

96.33

95.61

91.61

HRNet

98.22

95.55

95.43

95.49

91.58

B-FGC-Net

96.98

95.32

94.75

95.03

90.04

Segformer

98.31

95.78

95.66

95.72

91.99

TDNet

98.39

96.44

95.32

95.88

92.25

DSymFuser

98.30

95.79

95.58

95.68

91.93

GDGNet

97.95

96.21

95.51

95.86

92.15

SparseFormer

98.16

96.11

95.67

95.89

92.27

SegTDformer

98.51

96.52

95.90

96.21

92.85

  1. The highest value for each metric is marked as bold.
Back to article page