[G] Data Science

[113] Deep learning TEM image segmentation for automated microstructural analysis of FePt-C recording media

*Nikita Kulesh¹, Anton Bolyachkin¹, Ippei Suzuki¹, Yukiko Takahashi¹, Hossein Sepehri-Amin¹ (1. NIMS)

Keywords:machine learning、transmission electron microscopy、magnetic recording、convolutional neural networks、FePt

Pipeline for automatic processing of TEM images based on deep learning image segmentation was developed, applied to FePt-C samples, used for generating a dataset and building an optimization algorithm