Continuous Biometric Identity Verification System
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for secure access to online services often rely on one-time authentication, which is insufficient for preventing identity theft during prolonged web service sessions, as they do not continuously verify the user's identity without disrupting their activity.
Innovation Solution
A method and system that combines facial recognition with additional biometric techniques like voice and keystroke recognition to continuously verify a user's identity throughout their online session, using webcam photos and microphone audio, with periodic or random checks to ensure authenticity without interrupting the user's activity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If one-time authentication is used, then user access is granted quickly, but identity security during prolonged sessions deteriorates
Solution Approach 1:
The system performs continuous biometric verification throughout the online session by periodically capturing facial images and comparing them with the reference model. This continuous action maintains identity security without requiring the user to re-authenticate manually, thus resolving the contradiction between quick access and prolonged security.
Solution Approach 2:
The system implements periodic verification at defined intervals during the session, capturing facial images at specific time points rather than continuously. This periodic approach maintains security while minimizing user distraction and system resource consumption, balancing access efficiency with identity verification.
2Reliability
If continuous biometric verification is implemented, then identity security is improved, but user distraction and activity interruption worsen
Solution Approach 1:
The system performs verification passively by automatically capturing facial images using the webcam without requiring user initiation or active participation. The user simply continues their activity while the system independently performs biometric checks, eliminating user distraction while maintaining verification.
Solution Approach 2:
Verification occurs at periodic intervals rather than continuously, allowing users to perform their activities without constant interruption. The system captures images at defined time points, providing security while minimizing impact on user experience and activity flow.
3Measurement precision
If multiple biometric techniques are combined, then verification accuracy is improved, but system complexity worsens
Solution Approach 1:
The system combines facial recognition with additional biometric techniques such as voice recognition and keystroke pattern recognition. By merging multiple verification methods, the system achieves higher verification accuracy and confidence levels, resolving the contradiction between accuracy and the inherent complexity of implementing multiple techniques.
Solution Approach 2:
The system is designed to support multiple biometric verification methods within a unified framework. The same system architecture handles facial, voice, and keystroke verification, making the complexity manageable through standardized processing and reducing the overall system burden despite multiple verification capabilities.
Data Source
Figure 1
AI summary
A method for verifying the identity of a user of an online service, comprising: when a user is connected to an online service from a user terminal, establishing a connection with a biometric data collecting_module; downloading one application for taking photos with the webcam of the user terminal and at least one application for capturing audio or for extracting keystroke pattern; while a session with the online service is active: taking a photo of the user terminal; sending each photo and associated metadata to a management module; automatically extracting features of each face which appears in said photo; comparing said features with a biometrical model of a reference photo of the user; repeating the former steps, thus continuously verifying the identity of the user connected to the online service; using a second biometrical technique for verifying the user identity, said second biometric technique being either sound recognition or keystroke pattern analysis. If voice is detected or a keystroke pattern is extracted, voice/keystroke features are extracted and compared with a reference model. If the result of a comparison does not unequivocally match the person in the photo/audio clip/keystroke pattern with the registered user, either informing the service provider or sending the photo/audio clip for manual validation. Besides, if sound recognition is used and the detected sound is voice, another photo may be automatically taken and sent to the management module for features extraction and comparison with a biometrical model of a reference photo of the user. Similarly, if keystroke pattern recognition is used, another photo may be taken every time a keystroke pattern is detected.