{"title":"基于深度学习的ViT技术诊断人脸变形的渐进式识别","authors":"M.K. Mohamed Faizal, S. Geetha, A. Barveen","doi":"10.1109/ICNWC57852.2023.10127374","DOIUrl":null,"url":null,"abstract":"The face-morphing attack, which occurs in private, public, and governmental institutions, is one of the most well-known in today’s world. Face recognition systems tend to be vulnerable if the face images are manipulated with duplicate images. Manipulated images are combined with the original image so that the images look like legitimate ones. Several face recognition studies are being conducted to determine whether the face images are manipulated. Using the DL algorithm, the face image is trained to attain the original and morphed face images by recognizing the face images. DL algorithms determine the images by classifying whether they are morphs or not recognizable to humans. In this paper, the foremost emphasis is on diagnosing the face recognition from those face-morphed images using the different DL techniques. Different DL techniques are effectively compared, where the ViT transformer attains improved accuracy when compared to Resnet, RNN, and CNN, respectively. This paper provides an overview of the various deep learning algorithms for detecting those face recognition images that focus on challenges and issues in the facial datasets from Face Recognition Kaggle dataset with training and testing image dataset. It determines the higher contrast in image efficiency and the evaluation of the face recognized images with an improved image.","PeriodicalId":197525,"journal":{"name":"2023 International Conference on Networking and Communications (ICNWC)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2023-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Diagnosing Progressive Face Recognition from Face Morphing Using ViT Technique Through DL Approach\",\"authors\":\"M.K. Mohamed Faizal, S. Geetha, A. Barveen\",\"doi\":\"10.1109/ICNWC57852.2023.10127374\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"The face-morphing attack, which occurs in private, public, and governmental institutions, is one of the most well-known in today’s world. Face recognition systems tend to be vulnerable if the face images are manipulated with duplicate images. Manipulated images are combined with the original image so that the images look like legitimate ones. Several face recognition studies are being conducted to determine whether the face images are manipulated. Using the DL algorithm, the face image is trained to attain the original and morphed face images by recognizing the face images. DL algorithms determine the images by classifying whether they are morphs or not recognizable to humans. In this paper, the foremost emphasis is on diagnosing the face recognition from those face-morphed images using the different DL techniques. Different DL techniques are effectively compared, where the ViT transformer attains improved accuracy when compared to Resnet, RNN, and CNN, respectively. This paper provides an overview of the various deep learning algorithms for detecting those face recognition images that focus on challenges and issues in the facial datasets from Face Recognition Kaggle dataset with training and testing image dataset. It determines the higher contrast in image efficiency and the evaluation of the face recognized images with an improved image.\",\"PeriodicalId\":197525,\"journal\":{\"name\":\"2023 International Conference on Networking and Communications (ICNWC)\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2023-04-05\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2023 International Conference on Networking and Communications (ICNWC)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICNWC57852.2023.10127374\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2023 International Conference on Networking and Communications (ICNWC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICNWC57852.2023.10127374","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Diagnosing Progressive Face Recognition from Face Morphing Using ViT Technique Through DL Approach
The face-morphing attack, which occurs in private, public, and governmental institutions, is one of the most well-known in today’s world. Face recognition systems tend to be vulnerable if the face images are manipulated with duplicate images. Manipulated images are combined with the original image so that the images look like legitimate ones. Several face recognition studies are being conducted to determine whether the face images are manipulated. Using the DL algorithm, the face image is trained to attain the original and morphed face images by recognizing the face images. DL algorithms determine the images by classifying whether they are morphs or not recognizable to humans. In this paper, the foremost emphasis is on diagnosing the face recognition from those face-morphed images using the different DL techniques. Different DL techniques are effectively compared, where the ViT transformer attains improved accuracy when compared to Resnet, RNN, and CNN, respectively. This paper provides an overview of the various deep learning algorithms for detecting those face recognition images that focus on challenges and issues in the facial datasets from Face Recognition Kaggle dataset with training and testing image dataset. It determines the higher contrast in image efficiency and the evaluation of the face recognized images with an improved image.