the nmr sequence wavelet coefficients mat are input as cnn-resnet, and the corresponding nmr sequence wavelet coefficients mat are used as cnn-resnet18 output for denoising in python in python

main.py
import scipy.io as sio

# Load wavelet coefficients matrices
coeffs_input = sio.loadmat('input_coeffs.mat')
coeffs_output = sio.loadmat('output_coeffs.mat')

# Get the wavelet coefficients arrays
input_array = coeffs_input['coeffs']
output_array = coeffs_output['coeffs']
272 chars
10 lines

main.py
input_mean = input_array.mean()
input_std = input_array.std()
output_mean = output_array.mean()
output_std = output_array.std()

input_norm = (input_array - input_mean) / input_std
output_norm = (output_array - output_mean) / output_std
237 chars
8 lines

main.py
import numpy as np

input_tensor = np.reshape(input_norm, (1, input_norm.shape[0], input_norm.shape[1], 1))
output_tensor = np.reshape(output_norm, (1, output_norm.shape[0], output_norm.shape[1], 1))
200 chars
5 lines

main.py
from keras.models import Model
from keras.layers import Input, Conv2D, BatchNormalization, Activation
from keras.layers import Add, Conv2DTranspose, UpSampling2D

# Define the input shape
input_shape = input_tensor.shape[1:]

# Build the CNN-ResNet18 model
inputs = Input(shape=input_shape)
x = Conv2D(filters=64, kernel_size=3, padding='same')(inputs)

for i in range(16):
    x = residual_block(x, filters=64)

x = Conv2DTranspose(filters=1, kernel_size=3, padding='same', activation='linear')(x)

model = Model(inputs=inputs, outputs=x)
540 chars
18 lines

main.py
model.compile(optimizer='adam', loss='mse')
model.fit(input_tensor, output_tensor, epochs=100, batch_size=1)
109 chars
3 lines