No abstract

Cloud computing is a technology that has gained rapid popularity in recent years. It has enabled use of immense computational power in a scalable and cost-efficient manner. Deployment of biometric technology in government and commercial organizations has become a standard security practice. However, independent biometric systems tend to be computationally and financially expensive, especially when user enrollment is high. A feasible solution is to create a biometric system on the cloud which can be used ubiquitously as an authentication service. In this paper, we propose a first cancelable biometric framework based on deep learning on the cloud. We establish that cloud is a good solution for biometric systems where intensive computation, quick response times, and high accuracy is required. INDEX TERMS Biometric privacy, cloud computing, cancelable biometrics, deep learning.

This chapter outlines the current state of the art of Kinect sensor gait and activity authentication. It also focuses on emotional cues that could be observed from human body and posture. It presents a prototype of a system that combines recently developed behavioral gait and posture recognition methods for human emotion identification. A backbone of the system is Kinect sensor gait recognition, which explores the relationship between joint-relative angles and joint-relative distances through machine learning. The chapter then introduces a real-time gesture recognition system developed using Kinect sensor and trained with SVM classifier. Preliminary experimental results demonstrate accuracy and feasibility of using such systems in real-world scenarios. While gait and emotion from body movement has been researched in the context of standalone biometric security systems, they were never previously explored for physiotherapy rehabilitation and real-time patient feedback. The survey of recent progress and open problems in crucial areas of medical patient rehabilitation and rescue operations conclude this chapter.

The fundamental goal of a revocable biometric system is to defend a user’s biometrics from being compromised. This research explores the application of deep learning or Convolutional Neural Networks to multi-instance biometrics. Modality features are transformed into revocable templates through the application of random projection. During the user authentication phase, we employ Support Vector Machines, chosen over three other alternative classifiers after carrying out a comparative study. Comparison of the proposed method over other standard deep learning models and performance evaluation before and after revocability have also been discussed. Results demonstrate ability to improve identification accuracy and provide sound template security. The system was validated on three multi-instance iris and fingervein databases.

No abstract