Sunny M Lee¹ (), Steven K Shevell¹; ¹University of Chicago

Does feature binding to objects depend on higher-level object percepts? Percepts are mentally constructed representations that are visually experienced, such as hues, motions, and three-dimensional objects, not physical stimulus properties they derive from. Similarity among physical stimulus features, however, typically is proposed to drive feature binding to objects. While binding by physical similarity is a plausible pre-percept mechanism, does higher-level perceptual similarity also contribute to feature binding? Aim: Distinguish between two levels of visual neural representation to determine if a higher-level perceptual representation can drive the conscious experience of color-motion feature misbinding (the incorrect perceptual pairing of a physical object color and physical motion direction for peripheral stimuli, induced by the physical color-motion pairing in the central stimulus). Method: Observers viewed dot kinematograms comprised of two sets of colored dots, one red and one green, and with dots moving vertically upwards or downwards. While fixating centrally, observers reported the perceived motion direction in the periphery for one set of colored dots. Each set of dots had one of two motion types: (1) Planar linear motion in which all dots moved at the same constant vertical speed or (2) Structure From Motion (SFM) that had dots accelerating/decelerating vertically to evoke a 3D percept of a rotating cylinder. The two sets of dots otherwise had identical feature values (size, density, luminance) and were matched in average speed. Results/Conclusion: Color-motion perceptual misbinding was induced in the periphery for the majority of the presentation time when both sets had the same motion type (either both planar or both SFM). When one set was planar and the other SFM, however, misbinding was significantly reduced despite the preponderance of matched stimulus features for all dots. This indicates that misbinding can be facilitated by similar higher-level perceptual representations of motion.

Acknowledgements: Supported by NIH EY-026618