Multiple Level Crowding: Crowding at the Object Parts Level and at the Object Configural level

Area of Science:

• Visual perception
• Cognitive psychology
• Neuroscience

Background:

• Visual crowding impairs target identification in the presence of flankers.
• Current theories suggest crowding arises from early-stage feature integration.
• Recent evidence indicates crowding may also involve high-level object representations, particularly with faces.

Purpose of the Study:

• To investigate if visual crowding occurs at the object configural level, beyond crowding at the part level.
• To differentiate crowding effects using non-face objects with varying flanker similarity (parts vs. configuration).

Main Methods:

• Measuring target (square or diamond) identification accuracy at different visual eccentricities.
• Using flankers that matched either the target's parts or its overall configuration.
• Comparing crowding effects for part-based versus configuration-based flankers.

Main Results:

• Crowding effects were observed for both part-level and configuration-level flankers.
• Target identification accuracy decreased with increasing eccentricity for both flanker types.
• Crowding from object configurations was more potent and spatially extensive than crowding from object parts.

Conclusions:

• Visual crowding operates at both the part and configural representation levels of objects.
• Configural crowding is a significant factor in object recognition, influencing perceptual organization.
• These findings extend the understanding of crowding beyond low-level features to higher-level object representations.