{"title":"基于人脸深层特征和手势能量图像的多生物特征识别系统","authors":"Onur Can Kurban, T. Yıldırım, Ahmet Bilgic","doi":"10.1109/INISTA.2017.8001186","DOIUrl":null,"url":null,"abstract":"Nowadays, with the increasing use of biometric data, it is expected that systems work robustly and they can give successful results against difficult situations and forgery. In face recognition systems, variables such as direction of light, facial expression and reflection makes identification difficult. With biometric fusion, both safe and high performance results can be achieved. In this work, Eurocom Kinect Face dataset and BodyLogin Gesture Silhouettes dataset are used to create a virtual dataset and they were fused with score level. For face database, VGG Face deep learning model was used as feature extractor and energy imaging method was used for extracting gesture features. Afterwards the features reduced by principal component analysis and similarity scores were produced with standard deviation Euclidean distance. The results show that face recognition achieved a high performance with deep learning features under different light and expression conditions, however, multi-biometric results have reached higher genuine match rate (GMR) performance and lower false acceptance rate (FAR). As a result of this process, gesture energy imaging can be used for person recognition and for multi biometric data.","PeriodicalId":314687,"journal":{"name":"2017 IEEE International Conference on INnovations in Intelligent SysTems and Applications (INISTA)","volume":"113 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"23","resultStr":"{\"title\":\"A multi-biometric recognition system based on deep features of face and gesture energy image\",\"authors\":\"Onur Can Kurban, T. Yıldırım, Ahmet Bilgic\",\"doi\":\"10.1109/INISTA.2017.8001186\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Nowadays, with the increasing use of biometric data, it is expected that systems work robustly and they can give successful results against difficult situations and forgery. In face recognition systems, variables such as direction of light, facial expression and reflection makes identification difficult. With biometric fusion, both safe and high performance results can be achieved. In this work, Eurocom Kinect Face dataset and BodyLogin Gesture Silhouettes dataset are used to create a virtual dataset and they were fused with score level. For face database, VGG Face deep learning model was used as feature extractor and energy imaging method was used for extracting gesture features. Afterwards the features reduced by principal component analysis and similarity scores were produced with standard deviation Euclidean distance. The results show that face recognition achieved a high performance with deep learning features under different light and expression conditions, however, multi-biometric results have reached higher genuine match rate (GMR) performance and lower false acceptance rate (FAR). As a result of this process, gesture energy imaging can be used for person recognition and for multi biometric data.\",\"PeriodicalId\":314687,\"journal\":{\"name\":\"2017 IEEE International Conference on INnovations in Intelligent SysTems and Applications (INISTA)\",\"volume\":\"113 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"1900-01-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"23\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2017 IEEE International Conference on INnovations in Intelligent SysTems and Applications (INISTA)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/INISTA.2017.8001186\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2017 IEEE International Conference on INnovations in Intelligent SysTems and Applications (INISTA)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/INISTA.2017.8001186","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
A multi-biometric recognition system based on deep features of face and gesture energy image
Nowadays, with the increasing use of biometric data, it is expected that systems work robustly and they can give successful results against difficult situations and forgery. In face recognition systems, variables such as direction of light, facial expression and reflection makes identification difficult. With biometric fusion, both safe and high performance results can be achieved. In this work, Eurocom Kinect Face dataset and BodyLogin Gesture Silhouettes dataset are used to create a virtual dataset and they were fused with score level. For face database, VGG Face deep learning model was used as feature extractor and energy imaging method was used for extracting gesture features. Afterwards the features reduced by principal component analysis and similarity scores were produced with standard deviation Euclidean distance. The results show that face recognition achieved a high performance with deep learning features under different light and expression conditions, however, multi-biometric results have reached higher genuine match rate (GMR) performance and lower false acceptance rate (FAR). As a result of this process, gesture energy imaging can be used for person recognition and for multi biometric data.