Ailocker: Authenticated image locker for video

Citations

WEB OF SCIENCE

0
Citations

SCOPUS

0

초록

This paper proposes a new object-based authenticated encryption scheme with a constant sized signature. We combine the deep learning algorithm and some cryptographic techniques, such as symmetric key encryption and digital signatures, to satisfy the requirements mentioned above. Based on the deep learning algorithm, the objects are detected and each object unit is encrypted with a different key, thereby enabling access per object unit. The applied forward secure digital signature scheme guarantees not to forge a signature on the already captured image frames even if the device is hijacked. Experimental results show that the proposed scheme is practical in a real-time system due to the high performance of signature generation (10.5ms per frame) and a constant signature size overhead.

키워드

AuthenticationDigital signatureForward-secure signatureObject detectionPrivacySurveillanceAuthenticationData privacyDeep learningElectronic document identification systemsInteractive computer systemsLearning algorithmsObject detectionReal time systemsSpace surveillanceAuthenticated encryption schemeCryptographic techniquesDigital signature schemesForward-secureForward-secure signaturesObject basedSignature generationSymmetric key encryptionCryptography
제목
Ailocker: Authenticated image locker for video
저자
Jihye KimHankyung KoHyunok Oh
DOI
10.1145/3297280.3297594
발행일
2019-04
유형
Conference Paper
저널명
Proceedings of the ACM Symposium on Applied Computing
Part F147772
페이지
1508 ~ 1511