Purchase this article with an account.
Wesley Chaney, Alina Liberman, David Whitney; Serial dependence in perception of biological motion. Journal of Vision 2016;16(12):274. https://doi.org/10.1167/16.12.274.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
We experience the visual world as a continuous and stable environment of objects and events, despite rapidly changing retinal input due to eye movements and visual noise. One proposed mechanism for maintaining this apparent stability is serial dependence of visual perception: our current visual experience is biased toward what we have perceived in the recent past, facilitating information processing in an autocorrelated world. Recent studies have shown that perception of orientation and facial identity are biased toward previously seen stimuli (Fischer and Whitney, 2014; Liberman et al., 2004). We investigated whether this same perceptual bias occurred in the judgment of biological motion, specifically heading direction of point light walkers. Accurate perception of heading direction of point light walkers requires integration of information over time into a dynamic structural representation based on prior assumptions about how the local 2D information is generated by a familiar 3D form of a walker. The purpose of the current experiment is to extend previous investigation of serial dependence in static 2D images to the realm of dynamic representations. Subjects viewed a single walker presented in the periphery and followed by a mask. Subjects were then asked to adjust a second walker to match the direction of heading of the target walker. Subjects consistently made errors biased toward the direction of the target walker presented on the previous trial when the heading directions were similar across the two contiguous trials. Our results show that serial dependence can occur even for perception of stimuli that require integration of complex information over time and prior assumptions about the structure of complex objects in the world.
Meeting abstract presented at VSS 2016
This PDF is available to Subscribers Only