Scene Regularity Interacts With Individual Biases to Modulate Perceptual Stability

Abstract

Sensory input is inherently ambiguous but our brains achieve remarkable perceptual stability. Prior experience and knowledge of the statistical properties of the world are thought to play a key role in the stabilization process. Individual differences in responses to ambiguous input and biases toward one or the other interpretation could modulate the decision mechanism for perception. However, the role of perceptual bias and its interaction with stimulus spatial properties such as regularity and element density remain to be understood. To this end, we developed novel bi-stable moving visual stimuli in which perception could be parametrically manipulated between two possible mutually exclusive interpretations: transparently or coherently moving. We probed perceptual stability across three composite stimulus element density levels with normal or degraded regularity using a factorial design. We found that increased density led to the amplification of individual biases and consequently to a stabilization of one interpretation over the alternative. This effect was reduced for degraded regularity, demonstrating an interaction between density and regularity. To understand how prior knowledge could be used by the brain in this task, we compared the data with simulations coming from four different hierarchical models of causal inference. These models made different assumptions about the use of prior information by including conditional priors that either facilitated or inhibited motion direction integration. An architecture that included a prior inhibiting motion direction integration consistently outperformed the others. Our results support the hypothesis that direction integration based on sensory likelihoods maybe the default processing mode with conditional priors inhibiting integration employed in order to help motion segmentation and transparency perception.

Categorical judgments do not modify sensory representations in working memory

“…For example, in a task sequence where subjects first had to make a categorical judgment about a stimulus feature before providing an estimate of the feature value (motion direction [ 7 – 9 ], orientation [ 10 , 11 ] or numerosity [ 9 , 12 , 13 ]), the resulting estimates showed systematic biases in favor of their preceding category choices. Similar bias patterns have also been found in sequential tasks that did not explicitly involve categorical judgments [ 8 , 14 – 17 ] or estimates [ 18 , 19 ], suggesting that they are a wide-spread and general phenomenon in hierarchical perceptual inference [ 20 ]. A general theory proposes that these bias patterns emerge because inference is intrinsically a top-down process where low-level feature estimates are conditioned on the observer’s preceding, (explicit or implicit) high-level categorical judgment [ 10 , 16 , 20 – 22 ].…”

Categorical judgments do not modify sensory information in working memory

“…Similar bias patterns have also been found in sequential tasks that did not explicitly involve categorical judgments 13,14,8,15,16 or estimates 17,18 , suggesting that these biases are a wide-spread and general phenomenon in hierarchical perceptual inference 19 . A general theory proposes that these bias patterns emerge because inference is intrinsically a top-down process where low-level feature estimates are conditioned on the observer's preceding, (explicit or implicit) high-level categorical judgment 20,21,15 .…”

Resetting of Auditory and Visual Segregation Occurs After Transient Stimuli of the Same Modality

“…Behaviorally these two paradigms share a number of important features, notably that perception switches between a single coherent stream to segregated streams of information. Both paradigms also tend to follow a similar initial response pattern where the first perceptual phase is integrated for a relatively long time, followed by switching back-and-forth between shorter duration segregated and integrated percepts, supporting the hypothesis that the integrated percept serves as the default, or baseline, given stimulus parameters that elicit bistable perception (Hupé and Rubin, 2003;Pressnitzer and Hupé, 2006;Li et al, 2019). Additional hallmarks of spontaneous bistable perception, such as unpredictability from one perceptual phase to the next, and observance of logarithmic distribution of phase durations, are also present in both the ABA and moving plaid streaming paradigms (Pressnitzer and Hupé, 2006;Carter et al, 2014;Denham et al, 2018).…”