This notebook compares neural networks trained to denoise simulated electron-microscopy data via supervised and unsupervised learning. It also includes a comparison with linear shift-invariant denoising.
Electron Microscopy
This page contains code to reproduce the computational experiments in the tutorial Learn, Denoise and Discover: A Guide to Deep Denoising with an Application to Electron Microscopy , involving networks trained to denoise electron-microscopy data. The code is in the form of Jupyter notebooks that can be run in Google colab. These notebooks use pretrained models. The full code repository, available on GitHub, contains the models, as well as code to train them.
This notebook visualizes the strategy learned by a neural network to denoise real electron-microscopy data, by visualizing equivalent filters that are equal to the network gradient with respect to its input.
More code for denoising electron-microscopy data, as well as simulated data, can be found here. In addition, see the following links for the code and real data used in the paper Visualizing nanoparticle surface dynamics and instabilities enabled by deep denoising.