Table 7 Cascade R-CNN performance on the Augmented PCB Defect dataset.

From: Attentive context and semantic enhancement mechanism for printed circuit board defect detection with two-stage and multi-stage object detectors

Method

mAP

\(D_1\)

\(D_2\)

\(D_3\)

\(D_4\)

\(D_5\)

\(D_6\)

Cascade R-CNN w/Resnet50 (baseline8a)

98.0

98.4

96.7

98.1

97.5

98.2

99.5

Cascade R-CNN w/Resnet101 (baseline8b)

98.3

98.1

97.5

98.2

97.7

98.8

99.5

Cascade R-CNN w/Resnet50+ ACASEM

98.6

98.9

97.6

98.6

98.0

99.4

99.2

Cascade R-CNN w/Resnet 101+ ACASEM

98.5

99.0

97.8

98.9

97.4

99.1

99.1

  1. \(D_1\) stands for open-circuit, \(D_2\) : short, \(D_3\) : mousebite, \(D_4\) : spur, \(D_5\) : spurious_copper and \(D_6\) : missing_hole defects.
  2. Significant values are in bold.
Back to article page