Perceiving Illumination Inconsistencies in Scenes

Abstract

The human visual system is adept at detecting and encoding statistical regularities in its spatiotemporal environment. Here, we report an unexpected failure of this ability in the context of perceiving inconsistencies in illumination distributions across a scene. Prior work with arrays of objects all having uniform reflectance has shown that one inconsistently illuminated target can 'pop out' among a field of consistently illuminated objects (eg Enns and Rensink, 1990 Science 247 721 723; Sun and Perona, 1997 Perception 26 519-529). In these studies, the luminance pattern of the odd target could be interpreted as arising from either an inconsistent illumination or inconsistent pigmentation of the target. Either cue might explain the rapid detection. In contrast, we find that once the geometrical regularity of the previous displays is removed, the visual system is remarkably insensitive to illumination inconsistencies, both in experimental stimuli and in altered images of real scenes. Whether the target is interpreted as oddly illuminated or oddly pigmented, it is very difficult to find if the only cue is deviation from the regularity of illumination or reflectance. Our results allow us to draw inferences about how the visual system encodes illumination distributions across scenes. Specifically, they suggest that the visual system does not verify the global consistency of locally derived estimates of illumination direction.

The human visual system's assumption that light comes from above is weak

“…One caveat is that the lighting cues in our experiment were directly adjacent to the ambiguous disks. In real scenes lighting can vary from place to place, so perhaps the visual system only allows lighting cues to affect the perceived shape of immediately adjacent objects (23) and relies on the light-fromabove prior for interpreting large regions that have no local lighting cues. To test this possibility, we eliminated local lighting cues in the strong cue stimulus by showing the target disk on a flat surface and removing the two rods adjacent to the target disk (Fig.…”

The Influence of Cast Shadows on Visual Search

“…In particular, they are based on stimuli much more impoverished than what is generally encountered in the real world. Consequently, different behavior may apply when more realistic stimuli are usedöfor example interpretation may succeed for shadows on a ceiling, provided that a sufficient number of cues exist in the scene (eg Ostrovsky et al 2001).…”

“…Lighting direction is certainly one possibility: lighting in the natural world is usually from above (Lynch and Livingston 1995), and a similar regularity also exists in artificial environments. But this constraint does not appear to apply to all types of shadow perception (eg Ostrovsky et al 2001). Furthermore, other possibilities exist, such as the constraint that items must be on a ground plane (a plane viewed from above), or that they not be on a ceiling (a plane viewed from below).…”

Attention and Visual Memory in Visualization and Computer Graphics

“…Content may change prior to final publication. orientation [16], [17], [18], [19] length [20], [21] closure [16] size [22], [23] curvature [21] density [23] number [20], [24], [25] hue [23], [26], [27], [28], [29] luminance [21], [30], [31] intersections [16] terminators [16] 3D depth [32], [33] icker [34], [35], [36], [37], [38] direction of motion [33], [38], [39] velocity of motion [33], [38], [39], [40], [41] lighting direction [42], [43] Fig. 2.…”