Mitigating shoulder spoofing vulnerabilities in mobile payment systems: a security framework

Area of Science:

• Computer Science
• Cybersecurity
• Human-Computer Interaction

Background:

• Mobile payment systems face security risks like shoulder spoofing, where sensitive data is visually intercepted.
• Current security measures (PIN masking, screen dimming) are insufficient against environmental threats in public spaces.

Purpose of the Study:

• To introduce Gaze-Aware Threat Detection with Contextual Scene Analysis (GATCSA), a novel system for real-time mobile payment security.
• To address the limitations of existing security mechanisms by detecting and responding to visual threats.

Main Methods:

• Utilizing the front-facing camera and on-device computer vision for environmental monitoring during transactions.
• Implementing a risk evaluation module that analyzes proximity, gaze, and focus duration to classify threats.
• Processing data locally to ensure user privacy and efficient operation on mobile devices.

Main Results:

• Demonstrated high accuracy in identifying threats across diverse real-world conditions (lighting, crowd density, device orientation).
• Showcased effective user responsiveness through contextual alerts and actionable suggestions.
• Validated the system's ability to safeguard against unauthorized visual access.

Conclusions:

• GATCSA significantly advances mobile payment security by integrating gaze tracking and environmental awareness.
• The system enhances user trust and privacy by providing adaptive protection against shoulder spoofing.
• Local data processing ensures privacy and efficient performance on resource-constrained mobile devices.