Detecting distortions of peripherally presented letter stimuli under crowded conditions

Thomas S. A. Wallis, Saskia Tobias, Matthias Bethge, Felix A. Wichmann

Introduction

Visual crowding is a fundamental bottleneck in human vision, occurring when objects in the periphery become difficult to recognize due to surrounding clutter. While extensive research has explored how crowding affects the identification of letters (e.g., distinguishing an 'A' from an 'H'), this study investigates a more granular aspect of visual perception: the detection of geometric distortions in familiar shapes. By applying controlled spatial distortions to letters, the researchers aimed to probe the intermediate stages of visual processing where the brain encodes contours and features before high-level recognition occurs.

Methodology

The study employed psychophysical experiments with five observers using a 4-alternative forced choice (4AFC) task. Participants were presented with four letters at 8 degrees of retinal eccentricity and asked to identify which one was distorted. Two types of spatial distortion were tested:

  1. Bandpass Noise (BPN): Cartesian (x, y) distortions using filtered noise to create localized "wiggles" in letter strokes.
  2. Radial Frequency (RF): Polar coordinate distortions that modulate the radius of pixels according to a sinusoid, creating global, periodic deformations.
Stimulus generation process: (A) base letter dimensions, (B) bandpass noise pattern, and (C) the resulting distorted letter.

Experiments compared sensitivity in isolated conditions (single letter) versus crowded conditions (target letter surrounded by four undistorted flankers). The stimuli were based on Sloan letters (D, H, K, N as targets), and the researchers have made their raw data and analysis code available via Zenodo.

Key Insights and Results

Frequency Tuning and Crowding

The researchers found that sensitivity to BPN distortions is not uniform but "tuned" to specific spatial frequencies. For isolated letters, sensitivity peaked at approximately 6.5 cycles/degree (c/deg). However, under crowded conditions, this peak shifted to a higher frequency (~8.5 c/deg), suggesting that crowding selectively masks lower-frequency spatial information.

Matrix of distorted letter 'D' stimuli across varying distortion frequencies (horizontal) and amplitudes (vertical).

Cartesian vs. Polar Processing

Unlike BPN, RF distortions showed a linear trend where sensitivity improved as the radial frequency increased from 2 to 12 cycles. This divergence suggests that the visual system processes global polar deformations differently than local Cartesian ones.

The Complexity of Crowding

A significant finding was the failure of "pop-out" effects. Even when the target was the only distorted object among undistorted flankers, crowding severely hampered detection. Furthermore, if the flankers themselves were distorted, detection thresholds increased by a factor of 2.1x to 3.3x. This supports a "complexity" model of crowding, where increased feature density or clutter in the scene overwhelms the brain's ability to isolate the target's structure.

Conclusion and Clinical Implications

This research highlights that crowding is not just a failure of identification but a significant impairment in perceiving image structure. These findings have important implications for understanding metamorphopsia—distorted vision often caused by retinal diseases. Patients may fail to notice their own visual distortions in daily life because the cluttered environments of the real world induce crowding, effectively masking the geometric anomalies. Future work will aim to develop image-based similarity metrics to unify the perception of different distortion types across various letter geometries.

Summary updated

Abstract

Abstract When visual features in the periphery are close together they become difficult to recognise: something is present but it is unclear what. This is called “crowding”. Here we investigated sensitivity to features in highly familiar shapes (letters) by applying spatial distortions. In Experiment 1, observers detected which of four peripherally-presented (8 deg of retinal eccentricity) target letters was distorted (spatial 4AFC). The letters were presented either isolated or surrounded by four undistorted flanking letters, and distorted with one of two types of distortion at a range of distortion frequencies and amplitudes. The bandpass noise distortion (“BPN”) technique causes spatial distortions in cartesian space, whereas radial frequency distortion (“RF”) causes shifts in polar coordinates. Detecting distortions in target letters was more difficult in the presence of flanking letters, consistent with the effect of crowding. The BPN distortion type showed evidence of tuning, with sensitivity to distortions peaking at approximately 6.5 c/deg for un-flanked letters. The presence of flanking letters causes this peak to rise to approximately 8.5 c/deg. In contrast to the tuning observed for BPN distortions, RF distortion sensitivity increased as the radial frequency of distortion increased. In a series of follow-up experiments we found that sensitivity to distortions is reduced when flanking letters were also distorted, that this held when observers were required to report which target letter was undistorted, and that this held when flanker distortions were always detectable. The perception of geometric distortions in letter stimuli is impaired by visual crowding.
BibTeX
@inproceedings{wallis2017detecting,
  author = {Thomas S. A. Wallis and Saskia Tobias and Matthias Bethge and Felix A. Wichmann},
  title = {{Detecting distortions of peripherally presented letter stimuli under crowded conditions}},
  booktitle = {Attention},
  year = {2017},
  doi = {10.3758/s13414-016-1245-x},
}
Figure