Modeling and Exploiting the Time Course of Chromatic Adaptation for Display Power Optimizations in Virtual Reality Journal: TOG Volume: 44 6 212 12 DOI: 10.1145/3763294 1287 CCS: Computing methodologies Perception CCS: Computing methodologies Virtual reality
Ethan Chen, Sushant Kondguli, Carl Marshall, Yuhao Zhu · 2025
Read Modeling and Exploiting the Time Course of Chromatic Adaptation for Display Power Optimizations in Virtual Reality Journal: TOG Volume: 44 6 212 12 DOI: 10.1145/3763294 1287 CCS: Computing methodologies Perception CCS: Computing methodologies Virtual reality by Ethan Chen, Sushant Kondguli, Carl Marshall, Yuhao Zhu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-23489.txt.
Introduction
Virtual reality headsets drain power on bright displays, yet the eye adapts to slow color shifts without noticing. Ethan Chen and 3 co-authors exploit that trait of chromatic adaptation to save power on OLED screens. Their method needs no gaze tracking. A new test setup models how the viewer's adaptation state follows the scene's light color over time.
From that model they compute a shift path for the scene light that cuts power while staying inside a set budget for visible loss. Gradual shifts looked far better than the instant shifts of prior work. Paired with brightness dimming from earlier work, the method cut display power by about a third with no measured drop in perceived quality.
Catalog record
| CREATOR | Ethan Chen, Sushant Kondguli, Carl Marshall, Yuhao Zhu |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 83 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-23489
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Modeling and Exploiting the Time Course of Chromatic Adaptation for Display Power Optimizations in Virtual Reality Journal: TOG Volume: 44 6 212 12 DOI: 10.1145/3763294 1287 CCS: Computing methodologies Perception CCS: Computing methodologies Virtual reality Ethan Chen, Sushant Kondguli, Carl Marshall, Yuhao Zhu Abstract . We introduce a gaze-tracking–free method to reduce OLED display power consumption in VR with minimal perceptual impact. This technique exploits the time course of chromatic adaptation, the human visual system’s ability to maintain stable color perception under changing illumination. To that end, we propose a novel psychophysical paradigm that models how human adaptation state changes with the scene illuminant. We exploit this model to compute an optimal illuminant shift trajectory, controlling the rate and extent of illumination change, to reduce display power under a given perceptual loss budget. Our technique significantly improves the perceptual quality over prior work that applies illumination shifts instantaneously. Our technique can also be combined with prior work on luminance dimming to reduce display power by 31% with no statistical loss of perceptual quality. Figure 1 . Illustration of our gaze-tracking–free VR display power saving method exploiting the time course of human chromatic adaptation. (a): Our algorithm gradually shifts the illuminant in the virtual scene from the CIE Standard Illuminant D65 (average daylight) to a green-ish tint. This is done by carefully modulating the trajectory and the rate of the illuminant traversal to maximize display power saving while minimizing perceptual impact. (b): The plot shows the trajectory of the illuminant traversal applied in (a) in the CIE u^{\prime}v^{\prime} chromaticity space. The colors in the diagram represent the estimated display power consumptions under different illuminants; the power is normalized relative to D65, and clipped at 1. All the illuminants that yield power savings relative to D65 are colored green; the illumination shift in (a) leads to about 31% power saving. (c): To bound perceptual loss during the illuminant shift, we propose a novel psychophysical paradigm to model human adaptation state under illumination changes. The model predicts the user adaptation state a(t) , (i.e., one’s endogenous notion of neutral illuminant) given the actual illuminant A(t) in the scene. We bound the difference between the internal adaptation state and the actual illuminant to ensure perceptual quality during illuminant shifts. Credit for the original image used in Fig. 1a goes to Pedro Szekely ( Szekely 2019 ) . † † cc-license: by-sa 1. Introduction Virtual reality (VR) devices are severely power constrained. As the need for higher frame rate and higher resolution keeps increasing, the display contributes to an increasingly significant portion of the total power consumption ( Duinkharjav et al. 2022 ; Leng et al. 2019 ) , and is a prime candidate for power optimization. As long been established, OLED display power consumption is color-dependent ( Dong et al. 2009 ; Dash and Hu 2021 ; Chen et al. 2024 ; Duinkharjav et al. 2022 ) . This paper presents a new method to exploit the color dependency of OLED display power consumption. Our technique relies on chromatic adaptation, the process by which our visual system scales it sensitivity to maintain a stable color perception under variable lighting conditions ( Fairchild 2013 ) (Sec. 2 ). For example, when moving from warm, incandescent lighting indoors to cooler sunlight outdoors, our visual system adapts such that neutral objects, e.g., a white sheet of paper, always appear white-ish, even though the “objective” color of the paper in the colorimetric sense has changed. We exploit this phenomenon in virtual scenes. Given a reference scene rendered under an illuminant S , our idea is to render the same scene with a different illuminant T ; the new rendering accounts for the photoreceptor sensitivity change from S to T and, accordingly, shifts all the pixel colors such that the new rendering 1) appears to be perceptually equivalent to the reference rendering and 2) reduces the display power consumption. The pixel color shift can be implemented as a real-time shader with little computational cost. While conceptually simple, leveraging chromatic adaptation is challenging, because adaptation is not an instantaneous phenomenon — the human visual system can take several minutes to fully adapt to a change in illuminants ( Gupta et al. 2020 ; Fairchild and Reniff 1995 ) . The gradual nature of chromatic adaptation means that instant shifts in the illumination of VR environments are noticeable, as previously demonstrated ( Chen et al. 2024 ) . Our goal, thus, is to devise a principled strategy to gradually adapt the scene illuminant such that the perceptual impact is minimized, while power savings are maximized. This entails two tasks. First, we must identify “lucrative” illuminants that, if shifted to, would lead to significant power savings. We identify such illuminants using the chromatic adaptation theory and display power modeling (Sec. 4 ). Second, we must determine how the current illuminant should be shifted. To that end, we propose a novel psychophysical paradigm that can query the rate and bounds of a user’s adaptation state in real-time. Using observations from the psychophysical study, we build a statistical inference model using Maximum Likelihood Estimation that estimates how a user’s adaptation state changes over time, given the trajectory of the illuminant change (Sec. 5 ). Using the statistical model, we formulate a power optimization problem to calculate the illuminant shift trajectory that minimizes power consumption under a given perceptual loss budget. The optimal trajectory is then applied online during rendering time; this is implemented as a real-time fragment shader using little compute budget on a mobile VR GPU (Sec. 6 ). Through extensive subjective studies on a Meta Quest Pro, we demonstrate that our algorithm leads to significantly better perceptual quality compared to a previous chromatic-adaptation–based power saving method under the same power budget (Sec. 7 ). When combined with uniform luminance dimming, our method achieves 31% power saving with no statistical degradation in perceptual quality. We also discuss the limitations of our method (e.g., when the adaptation state has to be reset) and how it could be extended to accommodate more diverse VR content (Sec. 8 ). In summary, this paper makes the following contributions: • We analyze the power-saving opportunities under chromatic adaptation by mining a natural-scene dataset. • We propose a novel psychophysical paradigm and statistical inference method to probe and model human adaptation state under dynamic (time-varying) illumination. • Leveraging the computational model, we propose a new display-power saving algorithm that dynamically shifts the scene illuminant to minimize power consumption under a perceptual degradation bound. • Through subjective validation studies, we show that our algorithm saves about 31% display power with no statistical loss in perceptual quality. 2. Preliminaries on Chromatic Adaptation Throughout the day, we encounter a variety of lighting conditions, from incandescent bulbs to natural sunlight, yet our color perception remains relatively stable. This constancy arises partially because our visual system adapts to differing lighting conditions through a process known as chromatic adaptation. We discuss the key concepts necessary for the rest of the paper here, and leave the mathematical details to Supplementary Material A . von Kries 1902 ; von Kries 1905 hypothesized that (in modern interpretations) we adapt by scaling the spectral sensitivities of individual classes of