Securing Social Identity in Mobile Platforms: Technologies for Security, Privacy and Identity Management

Preface
Acknowledgments
Contents
About the Editors and Contributors
	Editors
	Contributors
Part I Mobile-Based Privacy & Security
	Shared Images and Camera Fingerprinting May Lead to PrivacyIssues
		1 Introduction
		2 Background
			2.1 Pre-processing
			2.2 Sensor Pattern Noise Extraction
			2.3 Classification and Clustering
			2.4 Similarity Measure for Camera Fingerprints
		3 Shared Image Analysis
			3.1 Smartphone Identification
			3.2 SPN-Based Image Clustering
			3.3 User Profile Linking
			3.4 Experimental Results
		4 Concluding Remarks
		5 Exercises
		References
	Presentation Attacks in Mobile and Continuous Behavioral Biometric Systems
		1 Introduction
			1.1 Biometric Performance
		2 Gait
		3 Keystroke Dynamics and Touch Gestures
		4 User-Device Interaction
		5 Open Challenges
		6 Summary
		References
	Personalized Data Minimization Assurance Using Bluetooth LowEnergy
		1 Introduction
		2 Related Work
		3 Overview of BLE Protocol
		4 Mobile Identification Data Profile
			4.1 Mobile Identification Data Service
			4.2 Service Characteristics
		5 Data Transferring Method
		6 Use Case Example
			6.1 Complete mID Data Request
				6.1.1 Offline Mode
				6.1.2 Online Mode
			6.2 Age Verification Data Request
		7 Conclusion
		References
Part II Mobile-Based Biometric Technologies
	On Designing a Forensic Toolkit for Rapid Detection of Factors that Impact Face Recognition Performance When Processing Large Scale Face Datasets
		1 Introduction
			1.1 Related Work
			1.2 Our Motivation and Contribution
		2 Methodology
			2.1 Databases
			2.2 Feature Extraction
			2.3 Conventional Models for Classification
			2.4 Convolutional Neural Networks for Classification
		3 Experimental Results
			3.1 Classification Results
		4 Conclusions and Future Work
		References
	Classification of Soft Biometric Traits When Matching Near-Infrared Long-Range Face Images Against Their Visible Counterparts
		1 Introduction
			1.1 Goals and Contributions
		2 Background
		3 Methodology
			3.1 System Design Set Up and Database Collected
			3.2 Proposed CNN Network for Automatic Prediction of Demographic Information
				3.2.1 Training and Testing
			3.3 Face Identification Steps
				3.3.1 Pre-processing of Images
				3.3.2 Combination of Pre-processing Methods
				3.3.3 Face Matching
		4 Experimental Results
			4.1 Baseline Face Matching Across Different Scenarios
			4.2 Demographic Grouping from Proposed CNN Architecture
			4.3 Face Matching Result Without Demographic Information Using Proposed Pre-processing Method
			4.4 Face Matching Results Without Demographic Information Using Proposed Fusion Scheme
			4.5 Experimental Results with Usage of Demographic Information Using Proposed Pre-processing and Fusion Scheme
				4.5.1 Cross-Matching Results for Long Distances
		5 Conclusion
		References
	Quality and Match Performance Analysis of Band-Filtered Visible RGB Images
		1 Introduction
			1.1 Literature Review
				1.1.1 Image Quality Assessment
				1.1.2 Face Image Quality Assessment
				1.1.3 Hyperspectral Face Detection and Recognition
		2 Methodology
			2.1 Data Collection
				2.1.1 Collection Scenarios
				2.1.2 Image Post-processing
				2.1.3 Data Collection Demographics
			2.2 Matching Evaluation of Band-Filtered Facial Images
				2.2.1 Matching Experiment Details
				2.2.2 Matching Experiment Results
			2.3 Image Quality Assessment
				2.3.1 Image Quality Assessment Experiments
				2.3.2 Image Quality Assessment Results
		3 Observations and Conclusions
		References
	Unconstrained Face Recognition Using Cell Phone Devices: Faces in the Wild
		1 Introduction
			1.1 Cell Phone Image Capture
		2 Cell Phone Face Database Creation
		3 Face Detection
			3.1 Traditional Methods
			3.2 Face Detection with Pose Estimation
		4 Face Recognition
			4.1 Previous Methods
				4.1.1 Simple Experiment
				4.1.2 Same Distance Matching
			4.2 State-of-the-Art Methods
		5 Summary
		References
	Face Detection in MWIR Spectrum
		1 Introduction
			1.1 Related Work
		2 Methodological Approach
			2.1 Deep Learning Models and Feature extractors
			2.2 Further Experiments Using Faster R-CNN with ResNet-101
		3 Experimental Evaluation
			3.1 Dataset Description
			3.2 Initial Experiments to Compare Different Models
			3.3 Different Experiments Using Faster R-CNN Model with ResNet 101
		4 Discussions
		5 Conclusions and Future Work
		Appendix
		References
Part III Mobile-Based Active Authentication
	Mobile Active Authentication based on Multiple Biometric and Behavioral Patterns
		1 Introduction
		2 Related Works
		3 System Description
			3.1 Behavioral-Based Profiling Systems
			3.2 Biometric Systems
		4 Experiments
			4.1 Database
			4.2 Experimental Protocol
				4.2.1 One-Time Authentication
				4.2.2 Active Authentication
		5 Results and Discussion
			5.1 One-Time Authentication
			5.2 Active Authentication
			5.3 Temporal Dependency in Behavioral-Based Profiling Systems
		6 Conclusions and Future Work
		References
	Quickest Multiple User Active Authentication
		1 Introduction
		2 Related Work
		3 Proposed Method
			3.1 Quickest Change Detection (QCD)
				3.1.1 CumSum Algorithm
			3.2 Efficient Quickest Change Detection
			3.3 Multi-user Quickest Intruder Detection (MQID)
		4 Experimental Results
			4.1 Protocol and Metrics
			4.2 Methods
			4.3 Results
		5 Concluding Remarks
		References
	Iris Recognition on Mobile: Real-Time Feature Extraction and Matching in the Wild
		1 Introduction
		2 Problem Statement
		3 Related Work
		4 Iris Feature Extraction and Matching
			4.1 Recognition Pipeline
			4.2 Low-Level Feature Representation
			4.3 Deep Feature Representation
			4.4 Matching Score Calculation
			4.5 Weighted Loss
			4.6 Multi-instance Iris Fusion
		5 Experimental Results
			5.1 Image-to-Image Evaluation
				5.1.1 Dataset Description
				5.1.2 Training
				5.1.3 Performance Evaluation
			5.2 Video-to-Video Evaluation
				5.2.1 Dataset Description
				5.2.2 Testing Procedure
				5.2.3 Performance Evaluation
		6 Conclusion
		References
	A Protocol for Decentralized Biometric-Based Self-Sovereign Identity Ecosystem
		1 Introduction
		2 Self-Sovereign Identity Ecosystem
		3 IEEE Biometric Open Protocol Standard (BOPS) Storage Model
			3.1 Server-Centric Approach
			3.2 Mobile-Centric Approach
			3.3 BOPS Distributed Storage Approach
		4 The Horcrux Protocol
		5 Summary
		References
	Towards Wider Adoption of Continuous Authentication on Mobile Devices
		1 Introduction
			1.1 User Acceptance
			1.2 Resource Consumption
		2 Related Work
			2.1 Continuous Authentication
			2.2 User Perceptions
			2.3 Resource Consumption
		3 User Perceptions
			3.1 Methodology: Survey Design
			3.2 Results
				3.2.1 Response Demographics
				3.2.2 Security Awareness
				3.2.3 Mobile Device OS
				3.2.4 Current Screen Lock Method (H6)
				3.2.5 Levels of Security (H7)
				3.2.6 Biometrics
				3.2.7 Continuous Authentication
			3.3 Discussion
			3.4 Limitations
		4 Resource Consumption
			4.1 Methodology
				4.1.1 Biometric Algorithms
				4.1.2 Datasets
				4.1.3 Resource Profile Curves (RPC)
				4.1.4 Measuring Energy Consumption (EC)
				4.1.5 Measuring Memory Consumption
			4.2 Results
				4.2.1 Calculating Energy Consumption
				4.2.2 Accuracy vs Energy Consumption Profile
				4.2.3 Accuracy vs Memory Consumption Profile
			4.3 Discussion
			4.4 Limitations
		5 Conclusions and Future Work
		References