Table 15 F4C AUG300 evaluation after LR changes.

From: Advanced deep learning modeling to enhance detection of defective photovoltaic cells in electroluminescence images

CNN models

Matrices

Precision

Recall

F1-Score

class 1

class 2

class 1

class 2

class 1

class 2

densenet169

0.9306

0.8424

0.928

0.8476

0.9293

0.845

densenet201

0.9371

0.8114

0.9086

0.8659

0.9226

0.8378

resnet101

0.9341

0.8011

0.903

0.8598

0.9183

0.8294

resnet152

0.9412

0.8512

0.9307

0.872

0.9359

0.8614

senet154

0.8854

0.7045

0.856

0.7561

0.8704

0.7294

vgg16

0.8943

0.8013

0.9141

0.7622

0.9041

0.7813

vgg19

0.9

0.869

0.9474

0.7683

0.9231

0.8155

Mean

0.9175

0.8116

0.9125

0.8188

0.9148

0.8143

Max

0.9412

0.869

0.9474

0.872

0.9359

0.8614

Min

0.8854

0.7045

0.856

0.7561

0.8704

0.7294

Back to article page