The relatively small decline in orientation acuity as stimulus size decreases

Area of Science:

• Visual neuroscience
• Computational vision
• Psychophysics

Background:

• Orientation acuity is crucial for spatial vision.
• Previous research indicates stimulus size affects visual performance.
• Existing models of spatial vision often employ feed-forward filter architectures.

Purpose of the Study:

• To investigate the effect of stimulus size on orientation acuity.
• To evaluate the predictive power of a feed-forward filter model for orientation acuity.
• To explore the role of primary visual cortex (V1) neuronal responses in explaining behavioral acuity.

Main Methods:

• Measured orientation acuity using circular patches of sinusoidal gratings of varying sizes.
• Compared empirical results with predictions from a Bayesian decision theory model using rectified Gabor filter outputs.
• Analyzed preliminary neuronal response data from V1 to refine the model.

Main Results:

• Orientation acuity was highest for the largest stimulus patches and decreased as patch size reduced.
• The standard feed-forward filter model quantitatively failed to predict thresholds for smaller stimuli.
• Incorporating V1 neuronal responses improved the model's ability to account for behavioral acuity.

Conclusions:

• The size-dependent nature of orientation acuity presents a challenge for simple feed-forward models of spatial vision.
• Feed-forward filter models with Bayesian decision theory do not fully capture the mechanisms underlying orientation acuity.
• Neuronal responses in V1 play a critical role in perceptual performance, suggesting limitations in current feed-forward computational models.