2018/03/19 by Mohammad Rashidi, Rashidi, Mohammad, Robert A. Wolkow +1 · 3 citations
Biochemistry, Genetics and Molecular Biology · Physics and Astronomy · #Advanced Electron Microscopy Techniques and Applications #Artificial intelligence #Characterization (materials science) #Computer science #Convolutional neural network #Dangling bond #FOS: Physical sciences #Force Microscopy Techniques and Applications #Materials science #Mesoscale and Nanoscale Physics (cond-mat.mes-hall) #Microscopy #Nanoscopic scale #Nanotechnology #Optics #Optoelectronics #Physics #Scanning electron microscope #Scanning probe microscopy #Scanning tunneling microscope #Silicon #Surface (topology) #Surface and Thin Film Phenomena #cond-mat.mes-hall
paper · pdf · doi:10.48550/arxiv.1803.07059
published in arXiv (Cornell University) 12(6), 5185-5189 (Cornell University)
openalex publication_date 2018/03/19 · arxiv created 2018/03/23 · arxiv updated 2018/03/26 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
Atomic scale characterization and manipulation with scanning probe microscopy\nrely upon the use of an atomically sharp probe. Here we present automated\nmethods based on machine learning to automatically detect and recondition the\nquality of the probe of a scanning tunneling microscope. As a model system, we\nemploy these techniques on the technologically relevant hydrogen-terminated\nsilicon surface, training the network to recognize abnormalities in the\nappearance of surface dangling bonds. Of the machine learning methods tested, a\nconvolutional neural network yielded the greatest accuracy, achieving a\npositive identification of degraded tips in 97% of the test cases. By using\nmultiple points of comparison and majority voting, the accuracy of the method\nis improved beyond 99%. The methods described here can easily be generalized to\nother material systems and nanoscale imaging techniques.\n