Table 8 Comparison of results attained in the ablation study on the proposed CAUC model.

From: Enhancing the weed segmentation in diverse crop fields using computationally effective concatenated attention U-Net with convolutional block attention module

Ver. no

Model name

Size

NPM

Validation time

Test time

Observation

FS

MIoU

FS

MIoU

Performance

Computational complexity

MB

Million

(%)

(%)

(%)

(%)

V1

CAUC-wo-CBAM_AG_LCB

63.4

5.51

98.84

80.05

96.89

78.52

High

Very High

V2

CAUC-wo-CBAM_AG

4.1

0.281

96.27

75.26

94.74

73.12

Low

~ 93.53% less than V1

V3

CAUC-wo-AG

4.3

0.288

97.70

79.11

95.82

77.55

Improved over V2

Less than V1

Slightly increased from V2

V4

CAUC-wo-CBAM

5.26

0.370

97.69

79.21

95.74

77.65

Improved over V2

Less than V1

Slightly increased from V2 & V3

V5

Actual proposed CAUC model

5.6

0.377

99.05

81.02

96.92

80.2

Higher than all

Slightly increased from V3 & V4.