Journal article
Untrained Physically Informed Neural Network for Image Reconstruction of Magnetic Field Sources
AEE Dubois, DA Broadway, A Stark, MA Tschudin, AJ Healey, SD Huber, JP Tetienne, E Greplova, P Maletinsky
Physical Review Applied | Published : 2022
Abstract
The prediction of measurement outcomes from an underlying structure often follows directly from fundamental physical principles. However, a fundamental challenge is posed when trying to solve the inverse problem of inferring the underlying source configuration based on measurement data. A key difficulty arises from the fact that such reconstructions often involve ill-posed transformations and that they are prone to numerical artifacts. Here, we develop a numerically efficient method to tackle this inverse problem for the reconstruction of magnetization maps from measured magnetic stray-field images. Our method is based on neural networks with physically inferred loss functions to efficiently..
View full abstractRelated Projects (1)
Grants
Awarded by National Centre of Competence in Research Robotics
Funding Acknowledgements
We acknowledge financial support from the National Centre of Competence in Research (NCCR) Quantum Science and Technology (QSIT), a competence center funded by the Swiss National Science Foundation (SNF), through SNF Project No. 188521, and by the European Research Council (ERC) consolidator grant project QS2DM. This work was supported by the Australian Research Council (ARC) through Grants No. CE170100012 and No. FT200100073.We would like to thank Sharidya Rahman, Boqing Liu, and Yuerui Lu for fabrication of the CuCrP2S6 sample. The machine-learning architecture was designed by A.E.E.D. and D.A.B. with assistance from P.M., A.S., E.G., and S.D.H. Author M.A.T. took the measurements of the CrI3 sample for in-plane simulations and testing of the machine-learning code. A.J.H. and J.-P.T. took measurements of the CuCrP2S6 sample for testing the angle determination. A.E.E.D., D.A.B., and P.M. wrote the manuscript. All authors discussed the data and commented on the manuscript.