Low-level image statistics in natural scenes influence perceptual decision-making

Area of Science:

• Cognitive Neuroscience
• Visual Perception
• Computational Neuroscience

Background:

• Human interaction with the environment relies on sensory information processing.
• Previous research often used artificial stimuli, limiting real-world applicability.
• Natural scenes possess regularities that the brain may exploit for decision-making.

Purpose of the Study:

• To investigate how natural scene complexity influences perceptual decision-making.
• To quantify the interaction between scene complexity and information processing.
• To determine the roles of contrast energy (CE) and spatial coherence (SC) in this interaction.

Main Methods:

• Participants performed an animal detection task on scenes of varying complexity.
• Scene complexity was defined by natural scene statistics: contrast energy (CE) and spatial coherence (SC).
• Diffusion modelling was used to analyze information processing speed and response boundaries.

Main Results:

• Low-level scene complexity significantly affected information processing speed.
• Performance was optimal for scenes of intermediate complexity.
• Isolated manipulation of spatial coherence (SC) affected processing speed and response boundaries, while contrast energy (CE) alone had no significant effect.

Conclusions:

• Natural scene complexity, characterized by CE and SC, modulates perceptual decision-making.
• The brain appears to adjust decision-making strategies based on input complexity.
• Intermediate complexity may represent an optimal range for efficient visual processing and decision-making.