Table 6 Performance comparison with existing semantic segmentation models. The bold font indicates the best values.

From: Segmentation of plateau zokor mounds in alpine meadows from UAV images using an improved UNet network

Model

Precision(%)

Recall(%)

MIoU(%)

Accuracy(%)

F1-score(%)

DD-DA

60.95

66.75

63.35

98.93

63.72

Res_AUNet

51.00

54.71

53.74

98.94

52.79

Twins

67.38

41.89

66.93

99.05

51.66

M-DeepLabV3+

59.07

58.54

70.32

99.01

58.80

V2-HRNet

63.21

56.60

70.82

99.08

59.72

M-PSPNet

46.21

41.18

63.28

98.71

43.55

VDP_UNet

71.44

65.63

75.63

99.27

68.41

Back to article page