Calibration Challenges and Data Quality in Eye Tracking for Inclusive XR Therapies

Extended Reality (XR) technologies offer significant potential in psychological therapies, especially for individuals with intellectual disabilities, who may need external support to enhance mental imagery and attention. Also, eye tracking data, which nowadays can be easily collected in XR applications, provides valuable insights into users’ cognitive and emotional states, further enriching these therapies. This study explores the challenges of creating inclusive XR applications, focusing on the eye tracking calibration process, which proved difficult for some users with intellectual disabilities during cognitive training sessions. In this sense, a simplified calibration scene was evaluated to assess gaze data quality, alongside a correction algorithm designed to enhance accuracy. While the algorithm improved accuracy in some cases, it frequently overfitted and failed to generalize. Data from uncalibrated sessions showed significant inaccuracy and inconsistency. Preliminary findings suggest that calibration challenges are likely related to participants’ vision impairments rather than intellectual disabilities. Although a preliminary study, these results offer insights and motivates the research on developing inclusive XR applications and improving calibration techniques for users with diverse needs.