December 2022
Volume 22, Issue 14
Open Access
Vision Sciences Society Annual Meeting Abstract  |   December 2022
Perceived Humanness Bias in Additive Light Model Displays
Author Affiliations
  • Austin Erickson
    University of Central Florida
  • Gerd Bruder
    University of Central Florida
  • Gregory Welch
    University of Central Florida
  • Isaac Bynum
    Davidson College
  • Tabitha Peck
    Davidson College
  • Jessica Good
    Davidson College
Journal of Vision December 2022, Vol.22, 4201. doi:https://doi.org/10.1167/jov.22.14.4201
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Austin Erickson, Gerd Bruder, Gregory Welch, Isaac Bynum, Tabitha Peck, Jessica Good; Perceived Humanness Bias in Additive Light Model Displays. Journal of Vision 2022;22(14):4201. https://doi.org/10.1167/jov.22.14.4201.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Additive light model displays, such as optical see-through augmented reality displays, create imagery by adding light over a physical scene. While these types of displays are commonly used, they are limited in their ability to display dark, low-luminance colors. As a result of this, these displays cannot render the color black and other similar colors, and instead the resulting color is rendered as completely transparent. This optical limitation introduces perceptual problems, as virtual imagery with dark colors appears semi-transparent, while lighter colored imagery is more opaque. We generated an image set of virtual humans that captures the peculiarities of imagery shown on an additive display by performing a perceptual matching task between imagery shown on a Microsoft HoloLens and imagery shown on a flat panel display. We then used this image set to run an online user study to explore whether this optical limitation introduces bias in user perception of virtual humans of different skin colors. We evaluated virtual avatars and virtual humans at different opacity levels ranging from how they currently appear on the Microsoft HoloLens, to how they would appear on a display without transparency and color blending issues. Our results indicate that, regardless of skin tone, the perceived humanness of the virtual humans and avatars decreases with respect to opacity level. As a result of this, virtual humans with darker skin tones are perceived as less human compared to those with lighter skin tones. This result suggests that there may be an unintentional racial bias when using applications involving telepresence or virtual humans on additive light model displays. While optical and hardware solutions to this problem are likely years away, we emphasize that future work should investigate how some of these perceptual issues may be overcome via software-based methods.

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×