User Authentication Using Inner-Wrist Skin Prints: Feasibility and Performance Assessment with Off-the-Shelf...

Area of Science:

• Biometrics
• Human-Computer Interaction
• Wearable Technology

Background:

• Continuous and implicit user authentication is crucial for wearable devices.
• Existing biometric modalities face challenges in reliability and practical integration.
• Inner-wrist skin texture is an underexplored biometric characteristic.

Purpose of the Study:

• To evaluate the feasibility of inner-wrist skin texture for biometric verification.
• To assess performance using off-the-shelf fingerprint sensing and matching technology.
• To investigate the impact of wrist posture on verification accuracy.

Main Methods:

• Two experiments were conducted using a capacitive fingerprint sensor and an unmodified algorithm.
• Experiment 1: Baseline performance assessment with 33 participants (1768 trials).
• Experiment 2: Effect of wrist posture variation with 15 participants (3900 trials).

Main Results:

• Experiment 1 showed a false acceptance rate upper bound of 6.7 × 10^-5 and a false rejection rate of 2.93%.
• Experiment 2 yielded a higher false rejection rate of 3.52% with no clear angle-performance relationship.
• Inner-wrist skin texture was successfully captured and matched using fingerprint technology under controlled settings.

Conclusions:

• Inner-wrist skin texture biometrics are viable with current fingerprint technology in controlled environments.
• Limitations include the use of closed matching algorithms and fingerprint-specific sensors, hindering interpretability.
• Future work should focus on dedicated recognition methods and larger sensing areas for wrist skin print biometrics.