Scene Regularity Interacts With Individual Biases to Modulate Perceptual Stability.

Area of Science:

• Cognitive Neuroscience
• Computational Neuroscience
• Visual Perception

Background:

• Perceptual stability arises from the brain's interpretation of ambiguous sensory input, influenced by prior experience and world knowledge.
• Individual differences in perceptual bias may modulate how ambiguous stimuli are interpreted, but their interaction with stimulus properties is unclear.

Purpose of the Study:

• To investigate how perceptual bias interacts with stimulus spatial properties (density, regularity) to influence perceptual stability.
• To model the role of prior knowledge in resolving ambiguous visual motion perception using hierarchical causal inference models.

Main Methods:

• Developed novel bi-stable moving visual stimuli allowing parametric manipulation between transparent and coherent motion interpretations.
• Employed a factorial design to probe perceptual stability across varying element densities and regularity levels (normal vs. degraded).
• Compared empirical data with simulations from four hierarchical causal inference models with different prior assumptions.

Main Results:

• Increased stimulus element density amplified individual biases, leading to enhanced perceptual stability for one interpretation.
• The effect of density on bias amplification was reduced under degraded regularity, indicating an interaction between these properties.
• A hierarchical model incorporating priors that inhibit motion direction integration best explained the observed data.

Conclusions:

• Perceptual bias and stimulus properties interact to stabilize perception of ambiguous visual motion.
• Conditional priors that inhibit motion direction integration appear crucial for resolving motion transparency and segmentation.
• This suggests a default mode of sensory likelihood integration modulated by inhibitory priors for complex motion perception.