Joint infrared pupil images and near-corneal-plane spectral light exposure data under natural conditions.

Factors influencing pupil size can be experimentally controlled or held constant in laboratory experiments, but how the pupil reflects changes in retinal input from the visual environment in real-world viewing conditions has yet to be captured in a large, age-diverse sample. In this dataset, we address this research gap by collecting data in a hybrid field-laboratory experiment (N = 83, 43 female, age range 18-87 years) using a custom-built, wearable, video-based eye tracker with a spectroradiometer measuring spectral irradiance in the approximate corneal plane, resulting in 29,664 valid recorded spectral irradiance and eye image pairs along with 83 approximately 3-minute-long calibration videos. After an initial 3-minute calibration procedure, 10-minute dark-adaptation period and a 14-minute controlled laboratory light condition, participants moved in and between indoor and outdoor environments of varying spectral irradiance for ~25-35 minutes and performed a range of everyday tasks. This dataset may provide a basis for developing algorithms for pupil detection, processing, and prediction under natural conditions.