Principles of gamma synchrony predict figure-ground perception in texture stimuli

Area of Science:

• Neuroscience
• Computational Neuroscience
• Visual Perception

Background:

• Gamma synchrony is prevalent in the visual cortex but its role in perceptual grouping is debated.
• Observed inconsistencies in gamma frequency and synchrony dependence on stimulus features challenge existing theories.
• The theory of weakly coupled oscillators (TWCO) suggests stimulus dependence is crucial for synchrony-based grouping.

Purpose of the Study:

• To investigate if stimulus-dependent gamma synchrony contributes to perceptual grouping and visual learning.
• To test the predictions of the TWCO model in a texture segregation task.
• To explore the role of gamma synchrony in training-induced improvements in visual perception.

Main Methods:

• Conducted a texture segregation experiment with human observers.
• Manipulated factors like feature dissimilarity and physical distance influencing synchrony.
• Developed a computational model based on TWCO with a Hebbian learning rule.

Main Results:

• Human performance in texture segregation quantitatively matched TWCO predictions.
• The computational model successfully predicted human learning effects.
• Increased gamma synchrony in the model correlated with improved texture segregation after learning.

Conclusions:

• Stimulus-dependent gamma synchrony captures local image statistics and is integral to texture segregation.
• Gamma synchrony's inherent stimulus dependencies provide a plausible mechanism for perceptual grouping and visual scene segmentation.
• Visual experience enhances gamma synchrony, offering a mechanism for perceptual learning in visual tasks.