Abstract
Image-recognition and text-based human verification challenges can be solved or imitated by automated agents, while more difficult challenges can increase user friction. A camera-based human verification technique presents a visual challenge and captures eye behavior during user interaction with the challenge. A model extracts eye data during the user interaction with the challenge such as gaze coordinates, micro-saccadic patterns, pupil diameter changes, and task-evoked pupillary response information. A verification engine correlates the timing of the pupillary response with cognitive milestones during the challenge, such as cognitive resolution, and can compare the observed response with a general human baseline. The technique supports liveness verification and bot detection while allowing image-derived eye-behavior processing to remain local to the device.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Kushwaha, Vivek, "Human Verification Based on Gaze Dynamics During Verification Challenge", Technical Disclosure Commons, (November 09, 2026)
https://www.tdcommons.org/dpubs_series/12838