Abstract
Millions of internet users around the world recently challenged science by asking why a certain photo of a dress led different observers to have surprisingly different judgments about the color of the dress. We suggest that a simple explanation provides a complete account of why observers see the colours of "the dress" in fundamentally, apparently irreconcilable ways (http://lpp.psycho.univ-paris5.fr/feel/?page_id=929). The reason is that the photo allows two different implicit interpretations about the illumination of the dress. Because observers have different assumptions about the illumination in the photo they perceive different colors of the dress. We tested this idea in two experiments. In a first experiment we measured the perceived colours of the dress, and the estimated color of the illumination: we did this by using (1) a color naming and (2) a color adjustment procedure. Naming and adjustment results show that the perceived color of the dress is negatively correlated with the assumed illumination along the daylight axis. In a second experiment we attempted to steer how observers would see the colors of the dress by manipulating their assumptions prior to seeing the photo of the dress. Prior to showing observers the original photo, we presented them two kinds of images showing the dress with disambiguated illumination cues. Each of the two disambiguated images represents one of two possible interpretations of the illumination conditions in the original photo. Depending on which of the disambiguated images observers saw, they were inclined to see the colors of the dress in the original photo in one or the other way. These findings confirm the idea that the perceived colors of the dress depend on the assumptions about the illumination. The phenomenon illustrates the power of prior assumptions in perception.
Meeting abstract presented at VSS 2016