Figure 5
A sketch of the workflow for training a CNN to extract parameters from augmented GISAS data. The GISAS data shown are a simulation of ordered nanoparticles on a surface. Data augmentation is done with several noise sources (including Poisson noise) and cropping of data in the centre to account for the beam stop that is typically present in experimental data. Image reproduced with permission from Van Herck et al. (2021) under a Creative Commons Attribution 4.0 International License, https://creativecommons.org/licenses/by/4.0/. |