He Yanmei, Thorstenson Christopher A
Munsell Color Science Laboratory, Rochester Institute of Technology, Rochester, NY, USA.
https://orcid.org/0009-0002-7927-7936.
J Vis. 2025 Aug 1;25(10):16. doi: 10.1167/jov.25.10.16.
Augmented reality (AR) aims to combine elements of the surrounding environment with additional virtual content into a combined viewing scene. Displaying virtual human faces is a widespread practical application of AR technology, which can be challenging in optical see-through AR (OST-AR) because of limitations in its color reproduction. Specifically, OST-AR's additive optical blending introduces transparency and color-bleeding, which is exacerbated especially for faces having darker skin tones, and for brighter and more chromatic ambient environments. Given the increasing prevalence of social AR applications, it is essential to better understand how facial color reproduction is impacted by skin tone and ambient lighting in OST-AR. In this study, a psychophysical experiment was conducted to investigate how participants adjusted colorimetric dimensions of OST-AR-displayed faces to match the color of the same faces viewed on a conventional emissive display. These adjustments were made for faces having six different skin tones, while under different simulated ambient luminance ("low" vs. "high") and chromaticity (warm, neutral, cool). Additionally, participants rated their adjustments for overall appearance match and preference. The results indicate that the magnitude and specific dimensions of colorimetric adjustments needed to make matches varied across skin tones and ambient conditions. The current work is expected to facilitate virtual human face reproduction in AR applications and to foster more equitable and immersive extended reality environments.
增强现实（AR）旨在将周围环境的元素与额外的虚拟内容结合到一个组合的观看场景中。显示虚拟人脸是AR技术广泛的实际应用，在光学透视式AR（OST-AR）中，由于其色彩再现的局限性，这可能具有挑战性。具体而言，OST-AR的加法光学混合会引入透明度和颜色渗出，对于肤色较深的人脸以及亮度更高、色彩更丰富的环境光来说，这种情况会更加严重。鉴于社交AR应用的日益普及，更好地了解在OST-AR中肤色和环境光如何影响面部色彩再现至关重要。在本研究中，进行了一项心理物理学实验，以调查参与者如何调整OST-AR显示的人脸的色度维度，使其与在传统发光显示器上看到的同一张脸的颜色相匹配。对具有六种不同肤色的人脸进行了这些调整，同时处于不同的模拟环境亮度（“低”与“高”）和色度（暖、中性、冷）条件下。此外，参与者对他们为整体外观匹配和偏好所做的调整进行了评分。结果表明，为实现匹配所需的色度调整的幅度和具体维度因肤色和环境条件而异。当前的工作有望促进AR应用中的虚拟人脸再现，并营造更公平、更具沉浸感的扩展现实环境。
