MagNet: machine learning enhanced three-dimensional magnetic reconstruction

TL;DR

This paper introduces MagNet, a deep learning-based method that significantly improves 3D magnetic field reconstruction from electron tomography data, especially in the presence of missing wedge artifacts.

Contribution

The paper presents a novel U-shaped CNN model trained on simulated magnetic textures to enhance VFET reconstructions, reducing artifacts caused by missing wedge effects.

Findings

MagNet outperforms conventional VFET in reconstructing magnetic fields.

Reconstruction quality is significantly improved with MagNet.

Deep learning effectively mitigates missing wedge artifacts in magnetic tomography.

Abstract

Three-dimensional (3D) magnetic reconstruction is vital to the study of novel magnetic materials for 3D spintronics. Vector field electron tomography (VFET) is a major in house tool to achieve that. However, conventional VFET reconstruction exhibits significant artefacts due to the unavoidable presence of missing wedges. In this article, we propose a deep-learning enhanced VFET method to address this issue. A magnetic textures library is built by micromagnetic simulations. MagNet, an U-shaped convolutional neural network, is trained and tested with dataset generated from the library. We demonstrate that MagNet outperforms conventional VFET under missing wedge. Quality of reconstructed magnetic induction fields is significantly improved.