Masking, crowding, and grouping: Connecting low and mid-level vision

Area of Science:

• Vision science
• Perceptual psychology
• Computational neuroscience

Background:

• Understanding low- and mid-level vision requires a unified framework.
• Spatial interactions between elements influence visual perception.
• Masking, crowding, and grouping are key spatial interaction processes.

Purpose of the Study:

• To investigate commonalities and differences between masking, crowding, and grouping.
• To characterize the spatial extent and contrast sensitivity of these visual processes.
• To determine how these processes vary with retinal eccentricity and inter-element spacing.

Main Methods:

• Measured contrast thresholds for Gabor detection, discrimination, and contour integration.
• Utilized a stimulus grid of nine Gabor elements with varying inter-element spacing.
• Analyzed baseline contrast and spatial window stability across parafoveal and peripheral vision.

Main Results:

• Contrast thresholds were modulated by inter-element distance, with grouping showing an inverted effect compared to masking and crowding.
• Baseline contrasts for detection and discrimination were positively linked, while those for integration and discrimination were negatively linked.
• Masking and crowding spatial windows were uncorrelated across eccentricity but correlated across participants, suggesting observer-specific computations.

Conclusions:

• Visual spatial interaction processes (masking, crowding, grouping) exhibit distinct spatial modulation patterns.
• Shared individual differences in spatial window extent point to underlying contrast-processing units.
• These findings contribute to a unitary framework for low- and mid-level vision.