Geometrically restricted image descriptors: A method to capture the appearance of shape

Area of Science:

• Visual perception
• Cognitive psychology
• Computational neuroscience

Background:

• Shape perception is variable and influenced by presentation conditions like visual field location.
• Existing methods often measure identification performance, failing to capture detailed shape appearance.
• Quantifying shape appearance is challenging, limiting our understanding of visual processing.

Purpose of the Study:

• Introduce Geometrically Restricted Image Descriptors (GRIDs) as a novel method to investigate shape appearance.
• Quantify how shape appearance is altered under specific viewing conditions.
• Analyze fine-grained details of shape perception beyond identification accuracy.

Main Methods:

• Developed the GRID paradigm using line elements on a grid to represent shapes.
• Observers recreated target shapes on a response grid, capturing perceived appearance.
• Presented stimuli at 10° eccentricity with gaze-contingent presentation to control eye movements.
• Analyzed data by quantifying differences between target and response shapes, including accuracy and error types.

Main Results:

• Demonstrated that GRIDs can effectively capture detailed shape appearance.
• Showed that fine-grained analysis of stimulus parts provides richer appearance quantifications than standard performance measures.
• Identified specific error types and element discriminability patterns in shape perception.

Conclusions:

• GRIDs offer a powerful and effective tool for investigating the nuances of shape appearance.
• This method provides a quantitative approach to understanding how visual stimuli are perceived.
• GRIDs advance the study of visual perception by detailing appearance beyond task performance.