Publication detail

Correction of AFM data artifacts using a convolutional neural network trained with synthetically generated data

KOCUR, V. HEGROVÁ, V. PATOČKA, M. NEUMAN, J. HEROUT, A.

Original Title

Correction of AFM data artifacts using a convolutional neural network trained with synthetically generated data

Type

journal article in Web of Science

Language

English

Original Abstract

AFM microscopy from its nature produces outputs with certain distortions, inaccuracies and errors given by its physical principle. These distortions are more or less well studied and documented. Based on the nature of the individual distortions, different reconstruction and compensation filters have been developed to post-process the scanned images. This article presents an approach based on machine learning - the involved convolutional neural network learns from pairs of distorted images and the ground truth image and then it is able to process pairs of images of interest and produce a filtered image with the artifacts removed or at least suppressed. What is important in our approach is that the neural network is trained purely on synthetic data generated by a simulator of the inputs, based on an analytical description of the physical phenomena causing the distortions. The generator produces training samples involving various combinations of the distortions. The resulting trained network seems to be able to autonomously recognize the distortions present in the testing image (no knowledge of the distortions or any other human knowledge is provided at the test time) and apply the appropriate corrections. The experimental results show that not only is the new approach better or at least on par with conventional post-processing methods, but more importantly, it does not require any operator's input and works completely autonomously. The source codes of the training set generator and of the convolutional neural net model are made public, as well as an evaluation dataset of real captured AFM images.

Keywords

Atomic force microscopy, Reconstruction by CNN, Machine learning for atomic force microscopy, Automatic image correction, Synthetic training data generation

Authors

KOCUR, V.; HEGROVÁ, V.; PATOČKA, M.; NEUMAN, J.; HEROUT, A.

Released

4. 1. 2023

ISBN

0304-3991

Periodical

Ultramicroscopy

Year of study

246

Number

1

State

Kingdom of the Netherlands

Pages from

113666

Pages to

113666

Pages count

16

URL

https://www.sciencedirect.com/science/article/pii/S0304399122001851?via%3Dihub

BibTex

@article{BUT183599,
  author="Viktor {Kocur} and Veronika {Hegrová} and Marek {Patočka} and Jan {Neuman} and Adam {Herout}",
  title="Correction of AFM data artifacts using a convolutional neural network trained with synthetically generated data",
  journal="Ultramicroscopy",
  year="2023",
  volume="246",
  number="1",
  pages="113666--113666",
  doi="10.1016/j.ultramic.2022.113666",
  issn="0304-3991",
  url="https://www.sciencedirect.com/science/article/pii/S0304399122001851?via%3Dihub"
}