/3D-CDI-NN

Neural Network for Coherent Diffraction Image Inversion

Primary LanguagePythonOtherNOASSERTION

3D-CDI-NN - Neural network for coherent diffraction image inversion

3D-CDI-NN is a deep neural network model plus automatic differentiation developed for retrieving phase information from 3D coherent diffraction images. The model is implemented using Tensorflow and the training dataset is generated using physics-based atomistic simulations. Custom codes are written to handle the resampling of diffraction images to oversampling ratios appropriate for the neural network model.

Comparing output from phase retrieval, 3D-CDI-NN prediction, and AD refined 3D-CDI-NN prediction

Reference:

"Rapid 3D nanoscale coherent imaging via physics-aware deep learning" Applied Physics Reviews 8, 021407 (2021) <https://doi.org/10.1063/5.0031486>

3D-CDI-NN is free software/open source, and is distributed under the BSD license. It contains third-party code, see below for the license information on third-party code:

Python <https://docs.python.org/3/license.html>
NumPy <https://github.com/numpy/numpy/blob/master/LICENSE.txt>
SciPy <https://scipy.org/scipylib/license.html>
scikit-image <https://github.com/scikit-image/scikit-image/blob/master/LICENSE.txt>
TensorFlow <https://github.com/tensorflow/tensorflow/blob/master/LICENSE>