Skip to main content

TableĀ 2 Network architecture of CSDAE

From: Automated chest screening based on a hybrid model of transfer learning and convolutional sparse denoising autoencoder

Layer (type)

Output shape

Param#

Input_1(InputLayer)

(None, 512, 512, 1)

0

Conv2d_1(Conv2D)

(None, 512, 512, 16)

160

Max_pooling2d_1(MaxPooling2D)

(None, 256, 256, 16)

0

Conv2d_2(Conv2D)

(None, 256, 256, 8)

1160

Max_pooling2d_2(MaxPooling2D)

(None, 128, 128, 8)

0

Conv2d_3(Conv2D)

(None, 128, 128, 8)

584

Max_pooling2d_3(MaxPooling2D)

(None, 64, 64, 8)

0

Conv2d_4(Conv2D)

(None, 64, 64, 8)

584

Up_sampling2d_1(UpSampling2D)

(None, 128, 128, 8)

0

Conv2d_5(Conv2D)

(None, 128, 128, 8)

584

Up_sampling2d_2(UpSampling2D)

(None, 256, 256, 8)

0

Conv2d_6(Conv2D)

(None, 256, 256, 8)

1168

Up_sampling2d_3(UpSampling2D)

(None, 512, 512, 16)

0

Conv2d_7(Conv2D)

(None, 512, 512, 1)

145