Performance of a deep convolutional neural network to classify crystal structures using selected area electron beam diffraction patterns containing lattice defect information.

Autor:	Jeong JM; Dept. of Materials Convergence and System Engineering, Changwon National University 20 Changwondaehak-ro Changwon-si Gyeongsangnam-do 51140 Republic of Korea woonglee@changwon.ac.kr., Ra M; LightVision Inc. 20 Seongsuil-ro 12-gil, Seongdong-gu Seoul 04793 Republic of Korea trizmaster@gmail.com., Jeong J; LightVision Inc. 20 Seongsuil-ro 12-gil, Seongdong-gu Seoul 04793 Republic of Korea trizmaster@gmail.com., Lee W; Dept. of Materials Convergence and System Engineering, Changwon National University 20 Changwondaehak-ro Changwon-si Gyeongsangnam-do 51140 Republic of Korea woonglee@changwon.ac.kr.; School of Materials Science and Engineering, Changwon National University 20 Changwondaehak-ro Changwon-si Gyeongsangnam-do 51140 Republic of Korea.
Jazyk:	angličtina
Zdroj:	RSC advances [RSC Adv] 2024 Jun 10; Vol. 14 (26), pp. 18489-18500. Date of Electronic Publication: 2024 Jun 10 (Print Publication: 2024).
DOI:	10.1039/d3ra08939h
Abstrakt:	A deep convolutional neural network (DCNN) architecture ResNet has been tested to verify its ability to handle selected area electron diffraction pattern (SADP) datasets carrying information on lattice defects including strains, thermal lattice vibrations, point defects, dislocations, and twin boundaries. The disordered states of the crystal lattices in the presence of these defects were predicted by ab initio molecular dynamics simulations, first principles geometry optimizations, and lattice manipulation operations in an effort to establish a possible dataset augmentation strategy for the improvement of classification performance of the ResNet. Using the disordered lattice information originating from the defects, test dataset SADPs were generated by simulating electron diffraction in transmission electron microscopy. The ResNet, pre-trained using SADPs from defect-free materials, showed decreasing but acceptable classification accuracies with increasing degrees of lattice disorder regarding the lattice vibrations and point defects. When tested using the diffraction patterns for strained lattices, the ResNet responded to the changing lattice symmetry when strain levels are relatively high suggesting that it has capability to discern different symmetries induced by large strains. However, the ResNet failed to recognize lattice structure when dislocations and twin boundaries were considered. It is suggested that DCNN architectures be trained over various scenarios including changes in the image feature characteristics in the diffraction patterns related to defects in future developments for improved general classification performances. Competing Interests: There are no conflicts to declare. (This journal is © The Royal Society of Chemistry.)
Databáze:	MEDLINE
Externí odkaz:	Zobrazit plný text záznamu