Skip to main content

Table 1 Detailed description of layers in DMF-Net

From: DMF-Net: a deep multi-level semantic fusion network for high-resolution chest CT and X-ray image de-noising

Operation Layer

No of Filters

Size of Each Filters

Dilation

Stride Value

Padding Value

Size of Output Image

Input Image

-

-

-

-

-

\(128 \times 128 \times 1\)

Convolution Layer

Convolution

64

\(3\times 3\times 1\)

1

\(1\times 1\)

\(1\times 1\)

\(128 \times 128 \times 64\)

(one)

ReLU

-

-

-

-

-

\(128 \times 128 \times 64\)

Convolution Layer

Convolution+BN

64

\(3\times 3\times 1\)

2

\(1\times 1\)

\(1\times 1\)

\(128 \times 128 \times 64\)

(one)

ReLU

-

-

-

-

-

\(128 \times 128 \times 64\)

Convolution Layer

Convolution+BN

64

\(3\times 3\times 1\)

1

\(1\times 1\)

\(1\times 1\)

\(128 \times 128 \times 64\)

(two)

ReLU

-

-

-

-

-

\(128 \times 128 \times 64\)

Convolution Layer

Convolution+BN

64

\(3\times 3\times 1\)

2

\(1\times 1\)

\(1\times 1\)

\(128 \times 128 \times 64\)

(one)

ReLU

-

-

-

-

-

\(128 \times 128 \times 64\)

Convolution Layer

Convolution+BN

64

\(3\times 3\times 1\)

1

\(1\times 1\)

\(1\times 1\)

\(128 \times 128 \times 64\)

(six)

ReLU

-

-

-

-

-

\(128 \times 128 \times 64\)

Convolution Layer

Convolution+BN

64

\(3\times 3\times 1\)

2

\(1\times 1\)

\(1\times 1\)

\(128 \times 128 \times 64\)

(one)

ReLU

-

-

-

-

-

\(128 \times 128 \times 64\)

Convolution Layer

Convolution+BN

64

\(3\times 3\times 1\)

1

\(1\times 1\)

\(1\times 1\)

\(128 \times 128 \times 64\)

(four)

ReLU

-

-

-

-

-

\(128 \times 128 \times 64\)

Convolution Layer

Convolution

1

\(3\times 3\times 1\)

1

\(1\times 1\)

\(1\times 1\)

\(128 \times 128 \times 1\)

(three)

-

-

-

-

-

-

\(128 \times 128 \times 1\)

 

Tanh

-

-

-

-

-

\(128 \times 128 \times 1\)

Convolution Layer

Convolution

1

\(3\times 3\times 1\)

1

\(1\times 1\)

\(1\times 1\)

\(128 \times 128 \times 1\)

(one)

ReLU

-

-

-

-

-

\(128 \times 128 \times 1\)

Convolution Layer

Convolution+BN

1

\(3\times 3\times 1\)

1

\(1\times 1\)

\(1\times 1\)

\(128 \times 128 \times 1\)

(one)

ReLU

-

-

-

-

-

\(128 \times 128 \times 1\)

Convolution Layer

Convolution+BN

64

\(3\times 3\times 1\)

1

\(1\times 1\)

\(1\times 1\)

\(128 \times 128 \times 64\)

(one)

ReLU

-

-

-

-

-

\(128 \times 128 \times 64\)

Convolution Layer

Convolution+BN

64

\(3\times 3\times 1\)

1

\(1\times 1\)

\(1\times 1\)

\(128 \times 128 \times 64\)

(three)

ReLU

-

-

-

-

-

\(128 \times 128 \times 64\)

Convolution Layer

Convolution+BN

1

\(3\times 3\times 1\)

1

\(1\times 1\)

\(1\times 1\)

\(128 \times 128 \times 1\)

(one)

ReLU

-

-

-

-

-

\(128 \times 128 \times 1\)