DATA LEARNING METHOD IN NEURAL NETWORK FOR IMAGE NOISE CANCELLATION

摘要

The present invention relates to a data learning method in a neural network for image denoising. In the present invention, the loss function for correcting the weight of the neural network is learned by the first epoch with the average absolute value error as the first learning rate, and then the loss function is changed to the structural similarity to the second epoch as the second learning rate. Disclosed is a data learning method in a neural network for image denoising, characterized in that it is trained. In the present invention, the weights are first learned to some extent through the L1 loss function, and then switched to the SSIM loss function to minimize the SSIM loss, so it can be performed at an appropriate operation speed and the learned weight A is the best performing neural network value. I was able to have confidence in my cognition.