Local stimulus disambiguation with global motion filters predicts adaptive surround modulation

Area of Science:

• Neuroscience
• Computational Neuroscience
• Visual Perception

Background:

• Humans effectively segment motion stimuli despite ambiguous local signals.
• Adaptive surround modulation (integrative/antagonistic modes) is key but not fully explained by local integration models.
• Existing models struggle to unify explanations for motion perception phenomena.

Purpose of the Study:

• Investigate local stimulus disambiguation using a distinct motion-processing model.
• Explain adaptive surround modulation and motion repulsion effects.
• Propose a unifying computational framework for motion perception.

Main Methods:

• Developed a novel motion-processing model based on global motion filters for velocity computation.
• Implemented inverse transformations for reconstructing local information from global signals.
• Introduced a novel filter within the architecture for local stimulus disambiguation.

Main Results:

• The global motion filter model successfully disambiguates local motion stimuli.
• The model demonstrates both integrative and antagonistic effects, matching human psychophysical data.
• A functional explanation for motion repulsion effects was provided.

Conclusions:

• A global motion filter approach offers a unifying explanation for motion perception phenomena.
• The novel embedded filter is crucial for achieving adaptive surround modulation.
• This model advances our understanding of visual motion processing and stimulus disambiguation.