Do Two Eyes Identify Better than One? Interpretable Binocular Dynamics for Virtual Reality Gaze Biometrics
Yiyu Zhang, Huijun Tang, Xiajing Yao · 2026
Read Do Two Eyes Identify Better than One? Interpretable Binocular Dynamics for Virtual Reality Gaze Biometrics by Yiyu Zhang, Huijun Tang, Xiajing Yao (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13514407.txt.
Description
Yiyu Zhang, Huijun Tang, and Xiajing Yao ask in 2026 whether the way two eyes move together can identify a person wearing a virtual reality headset. Earlier gaze biometrics leaned on one eye alone, leaving coordination between eyes largely untested. Using the GazeBaseVR dataset, the authors built features from eye misalignment, vergence movement, and fixation disparity, then checked them across sessions, tasks, and replay attacks.
Adding two-eye features to one-eye features raised first-guess identification across sessions, while the two-eye features alone carried some signal but not enough to stand alone. Performance held best when enrolment and test tasks matched, and dropped sharply when tasks changed. A neural sequence model also did better with true two-eye input than with a doubled one-eye feed. The authors position binocular movement as supporting evidence for headset logins, not a complete authenticator.
Catalog record
| CREATOR | Yiyu Zhang, Huijun Tang, Xiajing Yao |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43985 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |