GaitPrivacyON: Privacy-preserving mobile gait biometrics using unsupervised learning
نویسندگان
چکیده
Numerous studies in the literature have already shown potential of biometrics on mobile devices for authentication purposes. However, it has been that, learning processes associated to biometric systems might expose sensitive personal information about subjects. This study proposes GaitPrivacyON, a novel gait verification approach that provides accurate results while preserving subject. It comprises two modules: i) convolutional Autoencoders with shared weights transform attributes raw data, such as gender or activity being performed, into new privacy-preserving representation; and ii) system based combination Convolutional Neural Networks (CNNs) Recurrent (RNNs) Siamese architecture. The main advantage GaitPrivacyON is first module (convolutional Autoencoders) trained an unsupervised way, without specifying subject protect. Two experimental examinated: MotionSense MobiAct databases; OU-ISIR database. achieved suggest significantly improve privacy keeping user higher than 96.6% Area Under Curve (AUC). To best our knowledge, this considers methods way.
منابع مشابه
Privacy preserving biometrics using partially homomorphic encryption
We will look into privacy preserving biometrics using the example of a fingerprint reader and partially homomorphic encryption. Therefore we will cover the basics necessary to understand the discussed subject, partially homomorphic encryption and fingerprint based authentication, as well as showing a concrete protocol and its implications on performance and security of the system. While securit...
متن کاملSecure and Privacy-Preserving User Authentication Using Biometrics
Identity management lies in the field of Information Security, presenting numerous attractive research categories. Biometrics have been established as a new approach to mitigate the limitations and weaknesses of traditional access methods of passwords and tokens. However, biometrics introduce new security and privacy risks since they cannot be easily revoked. Due to the noisy nature of biometri...
متن کاملPrivacy Preserving Key Generation for Iris Biometrics
In this work we present a new technique for generating cryptographic keys out of iris textures implementing a key-generation scheme. In contrast to existing approaches to iris-biometric cryptosystems the proposed scheme does not store any biometric data, neither in raw nor in encrypted form, providing high secrecy in terms of template protection. The proposed approach is tested on a widely used...
متن کاملA Hybrid Deep Learning Architecture for Privacy-Preserving Mobile Analytics
The increasing quality of smartphone cameras and variety of photo editing applications, in addition to the rise in popularity of image-centric social media, have all led to a phenomenal growth in mobile-based photography. Advances in computer vision and machine learning techniques provide a large number of cloud-based services with the ability to provide content analysis, face recognition, and ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Pattern Recognition Letters
سال: 2022
ISSN: ['1872-7344', '0167-8655']
DOI: https://doi.org/10.1016/j.patrec.2022.07.015