New user authentication method based on eye-writing patterns identified from electrooculography for virtual reality...

Area of Science:

• Biomedical Engineering
• Human-Computer Interaction
• Virtual Reality Security

Background:

• Increasing demand for secure user authentication in virtual reality (VR) applications.
• Limitations of traditional VR authentication methods (e.g., hand controllers) including inconvenience and time consumption.
• Need for efficient and secure authentication for VR environments handling sensitive data.

Purpose of the Study:

• To propose and evaluate a novel user authentication method for VR based on eye-writing patterns.
• To leverage electrooculogram (EOG) signals for identifying unique eye-writing patterns for secure authentication.
• To assess the performance and effectiveness of the proposed EOG-based VR authentication system.

Main Methods:

• Recording electrooculogram (EOG) data from four locations around the eyes within a VR headset face-pad.
• Converting EOG data into a ten-dimensional similarity vector using dynamic time warping against ten pre-defined template patterns.
• Implementing a leave-one-subject-out cross-validation scheme with 19 participants to evaluate authentication accuracy.
• Defining a threshold for the similarity vector distance to determine successful authentication.

Main Results:

• The proposed eye-writing authentication method achieved an average accuracy of 97.74%.
• Demonstrated a low false accept rate of 1.31% and a false reject rate of 3.50%.
• The system showed effectiveness without requiring significant computational resources, suitable for edge devices.

Conclusions:

• The EOG-based eye-writing authentication method provides a secure and efficient solution for VR applications.
• This technique offers a promising alternative to conventional authentication methods, enhancing user privacy and data security in VR.
• The low computational burden makes it practical for implementation on edge devices in real-world VR scenarios.