ETRI-Knowledge Sharing Plaform

ENGLISH

성과물

논문 검색
구분 SCI
연도 ~ 키워드

상세정보

학술대회 Learning based Biometric Key Generation Method using CNN and RNN
Cited 3 time in scopus Download 2 time Share share facebook twitter linkedin kakaostory
저자
노종혁, 조상래, 진승헌
발행일
201807
출처
International Conference on Information Technology and Electrical Engineering (ICITEE) 2018, pp.136-139
출판사
IEEE
DOI
https://dx.doi.org/10.1109/ICITEED.2018.8534873
협약과제
18HH4600, 고신뢰 지능정보 서비스에서 휴먼(H)-인프라(I)-서비스(S)를 연결하는 Portal Device 보안 기술 개발, 조상래
초록
Studies on the biometric authentication and biometric key generation have been underway for a long time. Fundamentally, there is a disadvantage that it is difficult to obtain uniform biometrics due to noise, and issues of security and privacy are still mentioned, but high user convenience is an advantage that cannot be ignored. Recently, the results of existing researches on biometric key generation show very good results. However, since the algorithms presented in many studies are suited to the specific dataset, applying these algorithms to different datasets makes it difficult to achieve the good results mentioned in the paper. The reason is probably because most datasets are collected in one place with one camera. We wanted to present a key generation method that is not limited to datasets, and we came up with a training-based method for this. In this paper, we propose a method with the convolutional neural network (CNN) and the recurrent neural network (RNN) for cryptographic key generation from face biometrics. CNN is used to extract the feature vector from the face image, and RNN generates the key from the feature vector. In the registration process, the RNN is iteratively trained. Experimental results on the databases show that the proposed approach is effective in the biometric key generation. The results for mixed database also show good performance.
키워드
Biometric key, Biometrics, CNN, RNN
KSP 제안 키워드
Biometric authentication, Biometric key, Convolution neural network(CNN), Cryptographic Key Generation, Face Image, Face biometrics, Feature Vector, Long Time, Recurrent Neural Network(RNN), security and privacy