Khairul Khaizi Mohd Shariff, Megat Syahirul Amin Megat Ali, Ahmad Ihsan Mohd Yassin, Noor Ezan Abdullah, Ali Abd Al-Misreb, Aisyah Hartini Jahidin
{"title":"Small but Diverse SEM Image Dataset: Impact of Image Augmentation on the Performance of AlexNet","authors":"Khairul Khaizi Mohd Shariff, Megat Syahirul Amin Megat Ali, Ahmad Ihsan Mohd Yassin, Noor Ezan Abdullah, Ali Abd Al-Misreb, Aisyah Hartini Jahidin","doi":"10.18421/tem122-34","DOIUrl":null,"url":null,"abstract":"To this date, scanning electron microscope has produced among the most complex and diverse images at nanoscale resolution. The highly magnified images of backscattered electrons reflected from the surface of samples are non-uniformed, even for the same class of images. The study investigates the impact of having a small but diverse dataset on the performance of AlexNet. A total of 160 samples from EUDAT Collaborative Database Infrastructure is used for the study. Compared to the use of new non-augmented samples to increase the size of dataset, image augmentation has been significantly improved classification performance and generalization ability of the AlexNet.","PeriodicalId":45439,"journal":{"name":"TEM Journal-Technology Education Management Informatics","volume":"7 1","pages":"0"},"PeriodicalIF":0.6000,"publicationDate":"2023-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"TEM Journal-Technology Education Management Informatics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.18421/tem122-34","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"COMPUTER SCIENCE, INFORMATION SYSTEMS","Score":null,"Total":0}
引用次数: 0
Abstract
To this date, scanning electron microscope has produced among the most complex and diverse images at nanoscale resolution. The highly magnified images of backscattered electrons reflected from the surface of samples are non-uniformed, even for the same class of images. The study investigates the impact of having a small but diverse dataset on the performance of AlexNet. A total of 160 samples from EUDAT Collaborative Database Infrastructure is used for the study. Compared to the use of new non-augmented samples to increase the size of dataset, image augmentation has been significantly improved classification performance and generalization ability of the AlexNet.
期刊介绍：
TEM JOURNAL - Technology, Education, Management, Informatics Is a an Open Access, Double-blind peer reviewed journal that publishes articles of interdisciplinary sciences: • Technology, • Computer and informatics sciences, • Education, • Management
京公网安备 11010802042870号
京ICP备2023020795号-1
分享
求助内容：
应助结果提醒方式：
扫码关注我们
