{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,14]],"date-time":"2026-02-14T07:40:04Z","timestamp":1771054804985,"version":"3.50.1"},"reference-count":57,"publisher":"IOP Publishing","issue":"4","license":[{"start":{"date-parts":[[2025,10,13]],"date-time":"2025-10-13T00:00:00Z","timestamp":1760313600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2025,10,13]],"date-time":"2025-10-13T00:00:00Z","timestamp":1760313600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"funder":[{"name":"National Science and Technology Council","award":["Project NSTC 112-2221-E-006 -151 -MY3"],"award-info":[{"award-number":["Project NSTC 112-2221-E-006 -151 -MY3"]}]},{"name":"Google Research Scholar Award"}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Mach. Learn.: Sci. Technol."],"published-print":{"date-parts":[[2025,12,30]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>Scanning electron microscopy (SEM) provides high-resolution nanoscale imaging crucial for advanced materials characterization. Recent advancements in deep learning have significantly enhanced SEM analysis by automating the identification of nanoscale features. However, the reliability of these models remains insufficiently explored. We investigate two popular deep learning architectures, ResNet-50 and Swin Transformer, by systematically injecting faults into their weight parameters for SEM characterization. Our analysis demonstrates how performance degrades under varying fault scenarios, and also pinpoints which layers and bit positions exhibit heightened vulnerability. These findings provide insights for dependable, high-throughput automated materials characterization, thereby paving the way for more reliable deployment of SEM-based deep learning in industrial and research environments.<\/jats:p>","DOI":"10.1088\/2632-2153\/ae0ab1","type":"journal-article","created":{"date-parts":[[2025,9,23]],"date-time":"2025-09-23T22:54:33Z","timestamp":1758668073000},"page":"045008","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":1,"title":["Reliability of deep learning models for scanning electron microscopy analysis"],"prefix":"10.1088","volume":"6","author":[{"ORCID":"https:\/\/orcid.org\/0009-0004-4111-7480","authenticated-orcid":false,"given":"Chuen-Wun","family":"Pai","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0002-5944-4469","authenticated-orcid":false,"given":"Hung-Wei","family":"Hsueh","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-1258-3279","authenticated-orcid":true,"given":"Shu-Han","family":"Hsu","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"266","published-online":{"date-parts":[[2025,10,13]]},"reference":[{"key":"mlstae0ab1bib1","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6501\/abd96d","type":"journal-article","article-title":"Electron microscopy in semiconductor inspection","volume":"32","author":"Nakamae","year":"2021","journal-title":"Meas. Sci. Technol."},{"key":"mlstae0ab1bib2","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.progsolidstchem.2014.02.001","type":"journal-article","article-title":"Recent progress in scanning electron microscopy for the characterization of fine structural details of nano materials","volume":"42","author":"Suga","year":"2014","journal-title":"Prog. Solid State Chem."},{"key":"mlstae0ab1bib3","doi-asserted-by":"publisher","first-page":"973","DOI":"10.1039\/c3ja50026h","type":"journal-article","article-title":"Elemental mapping of microstructures by scanning electron microscopy-energy dispersive x-ray spectrometry (SEM-EDS): extraordinary advances with the silicon drift detector (SDD)","volume":"28","author":"Newbury","year":"2013","journal-title":"J. Anal. At. Spectrom."},{"key":"mlstae0ab1bib4","doi-asserted-by":"publisher","first-page":"20577","DOI":"10.1021\/acs.energyfuels.4c03783","type":"journal-article","article-title":"Analysis of pore types in lower cretaceous qingshankou shale influenced by electric heating","volume":"38","author":"Dong","year":"2024","journal-title":"Energy Fuels"},{"key":"mlstae0ab1bib5","doi-asserted-by":"publisher","DOI":"10.1088\/2632-2153\/abd614","type":"journal-article","article-title":"Deep learning in electron microscopy","volume":"2","author":"Ede","year":"2021","journal-title":"Mach. Learn. Sci. Technol."},{"key":"mlstae0ab1bib6","doi-asserted-by":"publisher","DOI":"10.1016\/j.mtnano.2020.100087","type":"journal-article","article-title":"Deep learning analysis on microscopic imaging in materials science","volume":"11","author":"Ge","year":"2020","journal-title":"Mater. Today Nano"},{"key":"mlstae0ab1bib7","first-page":"2886","type":"conference-proceedings","article-title":"Deep learning-based autonomous scanning electron microscope","author":"Jang","year":"2020"},{"key":"mlstae0ab1bib8","doi-asserted-by":"publisher","DOI":"10.1016\/j.ultramic.2020.113125","type":"journal-article","article-title":"Deep learning for scanning electron microscopy: synthetic data for the nanoparticles detection","volume":"219","author":"Kharin","year":"2020","journal-title":"Ultramicroscopy"},{"key":"mlstae0ab1bib9","doi-asserted-by":"publisher","first-page":"10761","DOI":"10.1039\/D2NR01029A","type":"journal-article","article-title":"A deep learning-based framework for automatic analysis of the nanoparticle morphology in SEM\/TEM images","volume":"14","author":"Sun","year":"2022","journal-title":"Nanoscale"},{"key":"mlstae0ab1bib10","doi-asserted-by":"publisher","DOI":"10.1088\/2632-2153\/add239","type":"journal-article","article-title":"Region versus query based instance segmentation models: application to the estimation of aggregated TiO2 particles size distribution measured by SEM","volume":"6","author":"Monchot","year":"2025","journal-title":"Mach. Learn. Sci. Technol."},{"key":"mlstae0ab1bib11","doi-asserted-by":"publisher","first-page":"88","DOI":"10.1038\/s41524-022-00767-x","type":"journal-article","article-title":"A deep learned nanowire segmentation model using synthetic data augmentation","volume":"8","author":"Lin","year":"2022","journal-title":"npj Comput. Mater."},{"key":"mlstae0ab1bib12","doi-asserted-by":"publisher","DOI":"10.1016\/j.ultramic.2024.113949","type":"journal-article","article-title":"Nano1D: an accurate computer vision software for analysis and segmentation of low-dimensional nanostructures","volume":"261","author":"Moradpur-Tari","year":"2024","journal-title":"Ultramicroscopy"},{"key":"mlstae0ab1bib13","first-page":"146","type":"conference-proceedings","article-title":"Edges detection of nanowires and adaptively denoising with deep convolutional neural network from SEM images","author":"Qu","year":"2020"},{"key":"mlstae0ab1bib14","doi-asserted-by":"publisher","DOI":"10.1016\/j.measurement.2024.114769","type":"journal-article","article-title":"Application of deep learning method for time reduction and precision improvement in displacement measurement during in-situ SEM tensile test","volume":"234","author":"Kim","year":"2024","journal-title":"Measurement"},{"key":"mlstae0ab1bib15","doi-asserted-by":"publisher","first-page":"216","DOI":"10.1117\/12.3014378","type":"conference-proceedings","article-title":"Advanced characterization of 2D materials using SEM image processing and machine learning","volume":"vol 12955","author":"Saib","year":"2024"},{"key":"mlstae0ab1bib16","first-page":"6884","type":"conference-proceedings","article-title":"Machine learning approach to estimate volumetric quantification of 2D SEM images of biofilms","author":"Abeyrathna","year":"2024"},{"key":"mlstae0ab1bib17","doi-asserted-by":"publisher","first-page":"46","DOI":"10.1038\/s41524-024-01226-5","type":"journal-article","article-title":"Deep learning for three-dimensional segmentation of electron microscopy images of complex ceramic materials","volume":"10","author":"Hirabayashi","year":"2024","journal-title":"npj Comput. Mater."},{"key":"mlstae0ab1bib18","doi-asserted-by":"publisher","first-page":"5877","DOI":"10.1038\/s41598-018-24330-1","type":"journal-article","article-title":"Super-resolution for asymmetric resolution of FIB-SEM 3D imaging using AI with deep learning","volume":"8","author":"Hagita","year":"2018","journal-title":"Sci. Rep."},{"key":"mlstae0ab1bib19","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-025-01763-z","type":"journal-article","article-title":"Automated cell structure extraction for 3D electron microscopy by deep learning","volume":"15","author":"Kousaka","year":"2025","journal-title":"Sci. Rep."},{"key":"mlstae0ab1bib20","doi-asserted-by":"publisher","first-page":"230212-1-230212","DOI":"10.29026\/oea.2024.230212","type":"journal-article","article-title":"Large-field objective lens for multi-wavelength microscopy at mesoscale and submicron resolution","volume":"7","author":"Xu","year":"2024","journal-title":"Opto-Electron. Adv."},{"key":"mlstae0ab1bib21","doi-asserted-by":"publisher","first-page":"924","DOI":"10.1109\/TMECH.2023.3323385","type":"journal-article","article-title":"Ultralarge-area stitchless scanning probe lithography and in situ characterization system using a compliant nanomanipulator","volume":"29","author":"Liu","year":"2024","journal-title":"IEEE\/ASME Trans. Mechatronics"},{"key":"mlstae0ab1bib22","doi-asserted-by":"publisher","DOI":"10.1002\/smll.202303838","type":"journal-article","article-title":"Ultra-large scale stitchless AFM: advancing nanoscale characterization and manipulation with zero stitching error and high throughput","volume":"20","author":"Liu","year":"2024","journal-title":"Small"},{"key":"mlstae0ab1bib23","doi-asserted-by":"publisher","DOI":"10.1117\/1.AP.6.4.046004","type":"journal-article","article-title":"Generative deep-learning-embedded asynchronous structured light for three-dimensional imaging","volume":"6","author":"Lu","year":"2024","journal-title":"Adv. Photon."},{"key":"mlstae0ab1bib24","doi-asserted-by":"publisher","first-page":"4958","DOI":"10.1109\/TII.2025.3547351","type":"journal-article","article-title":"A calibration and error evaluation method of a combined tracking-based vision measurement system for meter-scale components","volume":"21","author":"Jiang","year":"2025","journal-title":"IEEE Trans. Ind. Inform."},{"key":"mlstae0ab1bib25","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6560\/acf98f","type":"journal-article","article-title":"Deep learning for fast denoising filtering in ultrasound localization microscopy","volume":"68","author":"Yu","year":"2023","journal-title":"Phys. Med. Biol."},{"key":"mlstae0ab1bib26","doi-asserted-by":"publisher","first-page":"48895","DOI":"10.1364\/OE.544968","type":"journal-article","article-title":"Three-stage training strategy phase unwrapping method for high speckle noises","volume":"32","author":"Li","year":"2024","journal-title":"Opt. Express"},{"key":"mlstae0ab1bib27","doi-asserted-by":"publisher","first-page":"193","DOI":"10.1145\/2492101.1555372","type":"journal-article","article-title":"DRAM errors in the wild: a large-scale field study","volume":"37","author":"Schroeder","year":"2009","journal-title":"ACM SIGMETRICS Perform. Eval. Rev."},{"key":"mlstae0ab1bib28","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/2541228.2541236","type":"journal-article","article-title":"Modeling the impact of permanent faults in caches","volume":"10","author":"S\u00e1nchez","year":"2013","journal-title":"ACM Trans. Archit. Code Optim."},{"key":"mlstae0ab1bib29","article-title":"Silent data corruptions at scale","author":"Dixit","year":"2021","type":"preprint"},{"key":"mlstae0ab1bib30","doi-asserted-by":"publisher","first-page":"265","DOI":"10.1145\/1353536.1346315","type":"journal-article","article-title":"Understanding the propagation of hard errors to software and implications for resilient system design","volume":"43","author":"Li","year":"2008","journal-title":"ACM SIGPLAN Not."},{"key":"mlstae0ab1bib31","doi-asserted-by":"publisher","first-page":"14","DOI":"10.1109\/MM.2003.1225959","type":"journal-article","article-title":"Trends and challenges in VLSI circuit reliability","volume":"23","author":"Constantinescu","year":"2003","journal-title":"IEEE Micro"},{"key":"mlstae0ab1bib32","first-page":"1","type":"conference-proceedings","article-title":"What happens when circuits grow old: aging issues in CMOS design","author":"Sapatnekar","year":"2013"},{"key":"mlstae0ab1bib33","first-page":"283","type":"conference-proceedings","article-title":"BEC: bit-level static analysis for reliability against soft errors","author":"Ko","year":"2024"},{"key":"mlstae0ab1bib34","doi-asserted-by":"publisher","first-page":"258","DOI":"10.1109\/MDT.2005.69","type":"journal-article","article-title":"Soft errors in advanced computer systems","volume":"22","author":"Baumann","year":"2005","journal-title":"IEEE Des. Test Comput."},{"key":"mlstae0ab1bib35","first-page":"145","type":"conference-proceedings","article-title":"Software implemented detection and recovery of soft errors in a brake-by-wire system","author":"Skarin","year":"2008"},{"key":"mlstae0ab1bib36","first-page":"1","type":"conference-proceedings","article-title":"Soft error trends and mitigation techniques in memory devices","author":"Slayman","year":"2011"},{"key":"mlstae0ab1bib37","doi-asserted-by":"publisher","first-page":"305","DOI":"10.1109\/TDMR.2005.853449","type":"journal-article","article-title":"Radiation-induced soft errors in advanced semiconductor technologies","volume":"5","author":"Baumann","year":"2005","journal-title":"IEEE Trans. Device Mater. Reliab."},{"key":"mlstae0ab1bib38","article-title":"Discriminating between soft errors and hard errors in RAM","author":"Lavery","year":"2008","type":"other"},{"key":"mlstae0ab1bib39","doi-asserted-by":"publisher","first-page":"76","DOI":"10.1109\/MDAT.2019.2952336","type":"journal-article","article-title":"Are cnns reliable enough for critical applications? an exploratory study","volume":"37","author":"Neggaz","year":"2019","journal-title":"IEEE Des. Test"},{"key":"mlstae0ab1bib40","first-page":"1","type":"conference-proceedings","article-title":"Assessing convolutional neural networks reliability through statistical fault injections","author":"Ruospo","year":"2023"},{"key":"mlstae0ab1bib41","first-page":"1","type":"conference-proceedings","article-title":"The robustness of modern deep learning architectures against single event upset errors","author":"Arechiga","year":"2018"},{"key":"mlstae0ab1bib42","first-page":"298","type":"conference-proceedings","article-title":"Hardware faults that matter: understanding and estimating the safety impact of hardware faults on object detection DNNs","author":"Qutub","year":"2022"},{"key":"mlstae0ab1bib43","first-page":"1","type":"conference-proceedings","article-title":"Resiliency of automotive object detection networks on GPU architectures","author":"Lotfi","year":"2019"},{"key":"mlstae0ab1bib44","doi-asserted-by":"publisher","first-page":"e1873","DOI":"10.1002\/stvr.1873","type":"journal-article","article-title":"Investigating the impact of transient hardware faults on deep learning neural network inference","volume":"34","author":"Rahman","year":"2024","journal-title":"Softw. Test. Verif. Reliab."},{"key":"mlstae0ab1bib45","doi-asserted-by":"publisher","first-page":"555","DOI":"10.1109\/TVLSI.2023.3238907","type":"journal-article","article-title":"Exposing reliability degradation and mitigation in approximate DNNs under permanent faults","volume":"31","author":"Siddique","year":"2023","journal-title":"IEEE Trans. Very Large Scale Integr. (VLSI) Syst."},{"key":"mlstae0ab1bib46","first-page":"770","type":"conference-proceedings","article-title":"Deep residual learning for image recognition","author":"He","year":"2016"},{"key":"mlstae0ab1bib47","first-page":"10012","type":"conference-proceedings","article-title":"Swin transformer: hierarchical vision transformer using shifted windows","author":"Liu","year":"2021"},{"key":"mlstae0ab1bib48","doi-asserted-by":"publisher","first-page":"347","DOI":"10.1109\/TDMR.2010.2050692","type":"journal-article","article-title":"Theoretical foundation for upsets in CMOS circuits due to high-power electromagnetic interference","volume":"10","author":"Iliadis","year":"2010","journal-title":"IEEE Trans. Device Mater. Reliab."},{"key":"mlstae0ab1bib49","doi-asserted-by":"publisher","first-page":"10","DOI":"10.1145\/3084447","type":"journal-article","article-title":"Understanding reduced-voltage operation in modern DRAM devices: experimental characterization, analysis, and mechanisms","volume":"1","author":"Chang","year":"2017","journal-title":"Proc. ACM Meas. Anal. Comput. Syst."},{"key":"mlstae0ab1bib50","doi-asserted-by":"crossref","DOI":"10.1145\/2934583.2934589","type":"conference-proceedings","article-title":"Voltage noise induced DRAM soft error reduction technique for 3D-CPUs","author":"Lu","year":"2016"},{"key":"mlstae0ab1bib51","doi-asserted-by":"publisher","first-page":"1703","DOI":"10.1017\/S1431927615009290","type":"journal-article","article-title":"Determining the causes of scanning distortions in SEM and FIB","volume":"21","author":"Pluska","year":"2015","journal-title":"Microsc. Microanal."},{"key":"mlstae0ab1bib52","doi-asserted-by":"publisher","first-page":"89","DOI":"10.1111\/j.1365-2818.2006.01672.x","type":"journal-article","article-title":"Elimination of scanning electron microscopy image periodic distortions with digital signal-processing methods","volume":"224","author":"P\u0142uska","year":"2006","journal-title":"J. Microsc."},{"key":"mlstae0ab1bib53","first-page":"9","type":"journal-article","article-title":"Reproducible method for calibrating the backscattered electron signal for quantitative assessment of mineral content in bone","volume":"4","author":"Boyce","year":"1990","journal-title":"Scanning Microsc."},{"key":"mlstae0ab1bib54","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/sdata.2018.172","type":"journal-article","article-title":"The first annotated set of scanning electron microscopy images for nanoscience","volume":"5","author":"Aversa","year":"2018","journal-title":"Sci. Data"},{"key":"mlstae0ab1bib55","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-017-13565-z","type":"journal-article","article-title":"Neural network for nanoscience scanning electron microscope image recognition","volume":"7","author":"Modarres","year":"2017","journal-title":"Sci. Rep."},{"key":"mlstae0ab1bib56","first-page":"25","type":"conference-proceedings","article-title":"Pytorchfi: a runtime perturbation tool for dnns","author":"Mahmoud","year":"2020"},{"key":"mlstae0ab1bib57","type":"journal-article","article-title":"PyTorch library for CAM methods","author":"Gildenblat J and contributors","year":"2021","journal-title":"GitHub"}],"container-title":["Machine Learning: Science and Technology"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ae0ab1","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ae0ab1\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ae0ab1","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ae0ab1\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ae0ab1\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ae0ab1\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ae0ab1\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ae0ab1\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,13]],"date-time":"2025-10-13T07:33:09Z","timestamp":1760340789000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ae0ab1"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,13]]},"references-count":57,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2025,10,13]]},"published-print":{"date-parts":[[2025,12,30]]}},"URL":"https:\/\/doi.org\/10.1088\/2632-2153\/ae0ab1","relation":{},"ISSN":["2632-2153"],"issn-type":[{"value":"2632-2153","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,10,13]]},"assertion":[{"value":"Reliability of deep learning models for scanning electron microscopy analysis","name":"article_title","label":"Article Title"},{"value":"Machine Learning: Science and Technology","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2025 The Author(s). Published by IOP Publishing Ltd","name":"copyright_information","label":"Copyright Information"},{"value":"2025-06-04","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2025-09-23","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2025-10-13","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}