An investigation of perceptual experience in hue-rotated Altered Reality

Le résumé fourni par la source

The visual system adapts to simple chromatic transformations (e.g., red filters), but whether adaptation occurs for more complex color transformations such as a hue rotation (e.g., blue sky turns magenta) is an open question. We conducted two experiments that exposed participants to real-time hue rotation in passthrough altered reality (AR) whilst doing normal everyday activities. A hue rotation of 120 or 240 degrees in HSL space was applied to a Meta Quest 3 AR headset’s look up table. In Experiment 1 we measured unique hue settings (yellow and blue) and categorical color constancy (colored chips were sorted into categories under different illuminations) before and after 4 hours of adaptation to the altered reality. In Experiment 2 we measured unique yellow and blue before, after, and during each hour of a 4-hour AR adaptation period. All psychophysical tests were conducted outside the headset in a dark room where all the color cues were removed to reduce de-adaptation. We also recorded participants’ introspections whilst immersed in a hue-rotated environment. Data were analyzed by comparing bootstrapped means in task performance pre- and post- adaptation, and Bayesian ANOVAs were used to investigate effects of adaptation on color constancy and unique hue settings. We found that adaptation to the AR environment did not disrupt categorical color constancy (BF10(Exp1) <0.001) or settings of unique yellow and blue (BF10(Exp1) < 0.3; BF10(Exp2) < 0.01). Collectively, the results favor the null hypothesis that there is no adaptation to hue rotation, suggesting a limit to mechanisms of color adaptation. Directed content analysis of participants’ introspections indicated that hue rotation causes confusion in object recognition, aversive responses to atypical food colors, and elements of derealization. The study highlights the potential of AR as a tool for understanding color perception and cognition.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.