Table 2 Comparison results on the two change detection datasets.

From: Urban change detection of remote sensing images via deep-feature extraction

Models

Params

(M)

FLOPs

(G)

OSCD

Pre_c/Rec_c/Pre_nc/Rec_nc/F1/Kappa

CD_Data_GZ Pre_c/Rec_c/Pre_nc/Rec_nc/F1/Kappa

FC-Siam-conc

15.46

50.77

35.29/25.59/96.18/97.56/29.67/26.62

69.06/45.25/93.24/97.38/54.68/50.18

DTCDSCN

41.07

7.21

33.96/29.05/92.60/94.65/32.77/28.87

60.97/55.09/97.76/85.45/56.63/40.71

SNUNet

12.03

27.44

34.21/27.64/93.28/95.02/32.95/29.15

67.46/50.58/90.98/93.85/48.10/42.22

BIT

26.30

68.50

35.73/29.79/97.95/95.63/34.05/30.93

63.09/51.23/91.98/91.40/56.72/48.64

AMTNet

16.45

14.80

34.76/29.66/95.83/97.01/33.56/30.73

57.25/55.63/90.69/94.97/53.01/49.18

ScratchFormer

55.58

48.79

39.22/49.90/96.04/97.78/48.93/45.95

56.29/56.14/89.41/96.21/56.50/50.48

BiUnet-Dense

43.10

53.54

43.22/57.13/97.93/96.09/49.21/46.18

58.29/56.03/94.36/94.83/57.14/51.74