Figure Ground Segregation in Texture Stimuli

MATLAB code utilized for data acquisition in the psychophysics experiment described in the article:

Karimian, A., Roberts, M.J., De Weerd, P., & Senden, M. (n.d.). Adaptive Synchronization of Neural Gamma Oscillations Mediates Figure Ground Perception in Texture Stimuli. Manuscript submitted.

Abstract

Gamma synchrony is ubiquitous in visual cortex, but its role in perception has been questioned due to its volatility, depending on feature content of local image elements and their physical distance. Alternatively, these dependencies may shape synchronisation in meaningful ways. Using the theory of weakly coupled oscillators (TWCO), we demonstrate that stimulus dependence is crucial for gamma's role in perception. Two factors determine synchronisation among coupled oscillators: dissimilarity of oscillation frequencies and coupling strength. In early visual cortex these relate to local feature dissimilarity and physical distance between image elements, respectively. We manipulated these factors in a texture segregation experiment and found that human performance followed TWCO predictions qualitatively and quantitatively, as formalized in a computational model. Furthermore, learning-induced changes in model synchrony predicted changes in participants' performance over several sessions of performing the task. These results offer insights into the functional role of gamma synchrony for visual scene segmentation.

Experiment Overview

Eight participants engaged in a task to discriminate the orientation of a rectangular figure within a texture stimulus made up of Gabor annuli. This figure region was distinguishable from the background by its contrast distribution, manipulated through two independent variables: Contrast Heterogeneity and Grid Coarseness.

• Contrast Heterogeneity: Refers to the variance in contrasts of the Gabor annuli within the figure region. This experiment utilized five levels of contrast heterogeneity for the figure: 0.01, 0.2575, 0.505, 0.7525, and 1. The background contrast was always set to maximum heterogeneity (1).

• Grid Coarseness: Represents the spacing between Gabor annuli, affecting the texture's density. The experiment tested five levels of grid coarseness: 1, 1.125, 1.250, 1.375, and 1.5.

The study comprised 9 sessions, each containing multiple blocks of trials. Every block presented 25 distinct combinations of Contrast Heterogeneity and Grid Coarseness.

Stimuli, Tasks, and Procedure

• Each stimulus was a full-screen irregular grid of non-overlapping Gabor annuli on a grey background, with each annulus having a diameter of 0.7°, a spatial frequency of 5.7 cycles/degree, and a mean luminance of 60.76 Cd/m2.
• A rectangular figure differing in contrast distribution from its surrounding texture was embedded in the lower right quadrant for sessions 1-8 and moved to the upper left quadrant for session 9 (transfer session) to test learning transfer.
• Participants indicated the figure's orientation (horizontal or vertical) using arrow keys, within a two-alternative forced-choice setup. The stimulus remained visible for up to 1000 ms unless a response was made or fixation was lost.
• Correct responses were followed by a green fixation point (500 ms), incorrect by a red point.
• The experiment was conducted in a dimly lit room, using a chin and head-rest to maintain a constant eye-screen distance of 57 cm on a 19'' Samsung SyncMaster 940BF LCD monitor. Stimuli display and response recording were facilitated by Psychtoolbox-3 for MATLAB, with eye movements monitored by an Eyelink 1000 eye-tracker.

Procedure for Handling Aborted Trials

Trials where participants broke fixation were aborted and rescheduled randomly within the session.

Parametrization of the Experimental Setup

• Eye-screen distance: 57 cm
• Stimulus presentation time: 1000 ms
• Inter-trial interval: 900 ms
• Each session included 30 blocks, each with 25 trials.

Transfer Session

Session 9 tested the transferability of learned skills by relocating the figure to the upper left quadrant, with participants uninformed of the specific quadrant to expect the figure.

Ethics

Participants provided written consent after full briefing, in alignment with the Helsinki Declaration. The local Ethical Committee of the Faculty of Psychology and Neuroscience (ERCPN) approved all procedures. Participants were compensated for their involvement.

Requirements

• MATLAB 64-Bit (Version 3.0.14 or later)
• Psychtoolbox-3
• Eyelink 1000 eye-tracker (for fixation monitoring)

License

This project is open-sourced under the MIT License. See the LICENSE file for more details.