Does surface completion fail to support uncrowding?

Area of Science:

• Visual perception
• Computational neuroscience

Background:

• Visual crowding describes the impaired perception of a target due to surrounding elements.
• Existing low-level explanations like local pooling are insufficient due to the influence of global stimulus configurations.
• Grouping has been proposed as a key factor in explaining crowding effects.

Purpose of the Study:

• To re-evaluate the role of grouping in visual crowding.
• To address the findings of Moore and Zheng (2024) which questioned grouping's role based on ineffective occlusion stimuli.
• To demonstrate that the LAMINART neural network model can explain existing findings.

Main Methods:

• Analysis of stimuli used by Moore and Zheng (2024) to assess their effectiveness in inducing perceptual occlusion.
• Simulation of the LAMINART neural network model using the aforementioned stimuli.
• Comparison of model simulations with experimental results.

Main Results:

• The stimuli employed by Moore and Zheng (2024) failed to elicit the intended perceptual occlusion effects.
• Simulations using the LAMINART model successfully replicated the results reported by Moore and Zheng.
• The model's existing characteristics adequately explain the experimental outcomes without modification.

Conclusions:

• The conclusions drawn by Moore and Zheng (2024) regarding the role of grouping in crowding are not supported due to flawed stimuli.
• The LAMINART model, incorporating grouping principles, remains a viable explanation for visual crowding phenomena.
• Further research is needed with appropriate stimuli to fully elucidate the mechanisms of visual crowding.