A biologically-based computational model of visual cortex that overcomes the X-junction illusion.

Area of Science:

• Neuroscience
• Computational Vision
• Primate Visual Cortex

Background:

• Intrinsic terminators of moving bars provide clear motion direction, but extrinsic terminators at intersections create ambiguous signals.
• Distinguishing true motion from conflicting signals is crucial for visual perception.

Purpose of the Study:

• To propose a computational model explaining how primate visual cortex neurons differentiate intrinsic from extrinsic terminators.
• To investigate the role of form-motion interactions in resolving motion ambiguity.

Main Methods:

• Developed a two-stage computational model simulating primate motion processing.
• First stage modeled V1 complex and end-stopped neurons.
• Second stage modeled MT neurons receiving both motion signals and form information from V1 center-surround neurons.

Main Results:

• V1 end-stopped neurons could not differentiate terminator types.
• V1 center-surround neurons, responding to contrast, showed weaker inhibition at intrinsic terminators.
• This allowed MT neurons to preferentially enhance unambiguous motion signals from intrinsic terminators over ambiguous signals from extrinsic terminators.

Conclusions:

• Interactions between form (contrast) and motion information in V1 center-surround neurons enable MT neurons to resolve motion ambiguity.
• The model predicts that primate visual cortex can accurately determine the direction of moving bars, even with overlapping stimuli.