A cortical edge-integration model of object-based lightness computation that explains effects of spatial context and...

Area of Science:

• Visual perception
• Computational neuroscience
• Image processing

Background:

• Perceived surface reflectance (lightness) is modeled by integrating local luminance steps.
• Human vision uses specific strategies for integrating luminance, influenced by grouping and top-down factors.
• Previous models highlight edge integration and the impact of observer interpretation on lightness computation.

Purpose of the Study:

• To model the combined influences of grouping and attention on lightness computation.
• To propose a cortical mechanism integrating spatial selection and object-based network computations.
• To explain how border ownership influences neural gains for edge signals in lightness perception.

Main Methods:

• Developing a computational model of lightness perception.
• Incorporating top-down signals for spatial selection of regions of interest.
• Utilizing object-based network computations involving border-ownership neurons.

Main Results:

• Demonstrated a model where top-down signals select regions, and border ownership sets neural gains for edge integration.
• Showed that only edges surviving both selection and gain control stages contribute to lightness computation.
• Integrated findings with neurophysiological data from V1, V2, and V4.

Conclusions:

• The proposed model explains how grouping and attention jointly influence lightness perception via a cortical mechanism.
• The model aligns with neurophysiological evidence for hierarchical visual processing in lightness computation.
• This work provides a framework for understanding the neural basis of surface reflectance perception.