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

main.py
import numpy as np

wavelet_coeffs = np.load('path/to/wavelet/coefficients/matfile')
85 chars
4 lines

main.py
# Reshape the array to add a "channel" dimension
wavelet_coeffs = wavelet_coeffs.reshape((1, *wavelet_coeffs.shape))

# Normalize the data
wavelet_coeffs = (wavelet_coeffs - np.min(wavelet_coeffs)) / (np.max(wavelet_coeffs) - np.min(wavelet_coeffs))
250 chars
6 lines

main.py
import torch
import torchvision

# Load the pretrained model
model = torchvision.models.resnet18(pretrained=True)
114 chars
6 lines

main.py
# Convert the numpy array to a PyTorch tensor
wavelet_coeffs_tensor = torch.Tensor(wavelet_coeffs)
99 chars
3 lines

main.py
# Set the model to evaluation mode
model.eval()

# Pass the wavelet coefficients through the model
denoised_output = model(wavelet_coeffs_tensor)

# Convert the denoised output to a numpy array
denoised_output = denoised_output.detach().numpy()
245 chars
9 lines

main.py
# Save the denoised output as a mat file
np.save('path/to/denoised/wavelet/coefficients/matfile', denoised_output)
115 chars
3 lines