Abstract
Purpose and Background: How does the visual system resolve multiple ambiguous features’ neural representations? Is resolution of features’ representations facilitated when the features are integrated within a single object, compared to resolution of the same feature representations in separate objects? When two visual stimuli share one ambiguous feature (e.g., color ambiguity from chromatic rivalry in both members of a pair of discs, or orientation ambiguity in a pair of Necker cubes), both stimuli are seen as identical more often than chance (Babich and Standing, 1981; Slezak and Shevell, 2018). This occurs also with a pair of stimuli having two ambiguous features (orientation and color; Lange and Shevell, OSA FVM 2018). Experiments here test the following question: In one or two pairs of objects, is perceived identity for both of two ambiguous features greater when those two ambiguous features are integrated components of each object, compared to when they are distributed across separate objects? Methods: Four observers completed two sessions, viewing in one session a pair of Necker cubes presented in red-green chromatic interocular-switch rivalry (cISR; Christiansen, D’Antona, and Shevell, 2017), and in the other session a pair of non-rivalrous gray Necker cubes adjacent to red-green cISR discs. In both sessions, observers reported via button presses when both rivalrous colors appeared identical and also when both Necker cubes’ orientations appeared identical. The proportion of viewing time with identical color-and-cube-orientation percepts was compared between sessions. Results and Conclusions: For three of four observers, identical-color-and-orientation proportions were not significantly different between the two sessions (always p > 0.1, tested separately for each observer). For one observer, identical-color-and-orientation proportions were significantly lower for cubes presented in red-green cISR (p< 0.0001). Thus, there was no evidence to support the hypothesis that integration of different features in a single object facilitated resolution of those ambiguous features.
Acknowledgement: Supported by NIH EY-026618