Eye Biometrics Signal Analysis and Potential Applications in User Authentication and Affective Computing
Bibliographic record
Abstract
As the adoption of biometric technologies grows, there is an escalating worry regarding the potential for bypassing, disguising, or replicating biometric data. Established biometric methods, such as facial recognition and fingerprints, are susceptible to these forms of drawbacks, undermining their core objective, which is to use physical traits for reliable deployment. This thesis proposes utilizing eye biometric signals (eye movement and pupillometry) for identifying individuals and recognizing affective states. Eye biometric signals are critical biological signals that inherently incorporate liveliness verification and resist counterfeiting due to their distinctive nature. However, utilizing eye signals as a biometric is challenging because recordings from the same person can vary due to factors like background noise, physical exertion, and emotions. While variations due to noise and physical activity can be managed through proper pre-processing and feature extraction techniques, the influence of psychological factors on eye biometric signals is more complicated to handle. This thesis deals with this problem from an affective computing point of view. First, the thesis pinpoints the psychological conditions that can influence the eye signals which, consequently, threaten the precision of its deployment in various applications. It outlines experimental arrangements purposely designed to induce active and passive arousal as well as positive and negative valence. A data-driven model serves as the foundation for detecting emotional patterns after tailoring the model for eye biometric signals. The results demonstrate the identification of psychological/affective states that affect eye biometrics to a degree that may render biometric matching unreliable. Three application scenarios are examined: a) user authentication, b) affective/emotional state recognition, and c) Post-traumatic Stress Disorder (PTSD) detection. The developed models are tailored to suit each context, and the pros and cons of each setup are explored. Additionally, to improve the utility of these emerging modalities, a multimodal framework is introduced that combines facial expressions and pupillometry. Finally, this thesis introduces a study that investigates the existence of bias with eye biometric signals and identifies the different factors that affect modality performance. Overall, this thesis endeavors to develop the necessary algorithms and practical infrastructure for effectively employing eye biometric signals in real-world biometric systems.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.006 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".