[20a-PA1-1~11] 7 Beam Technology and Nanofabrication (Poster)

[20a-PA1-4] Convolutional Neural Network for electron diffraction pattern classification

〇Fumihiko Uesugi¹, Kazutaka Mitsuishi¹, Koji Kimoto¹, Masashi Ishii¹ (1.National Institute for Materiasls Science)

Keywords:electron diffraction pattern, Convolutional Neural Network, classification

A convolutional neural network (CNN) system is attracting attention with image classification and recognition.
Electron diffraction patterns (DP) have a lot of information from crystals but their analyses are boring even if experts. If the CNN system is effective for DP analyses, loads of analysts would be decreased.
So, we applied CNN system to classify crystals with DP and studied the effectiveness and the possibility. We will show the results and the future plans.