Repulsive Aftereffects of Visual Space

Abstract

Prolonged exposure to a sensory stimulus induces perceptual adaptation aftereffects. Traditionally, aftereffects are known to change the appearance of stimulus features, like contrast, color, or shape. However, shifts in the spatial position of objects have also been observed to follow adaptation. Here, I demonstrate that visual adaptation produced by different adapter stimuli generates a bi-directional spatial repulsion. Observers had to judge the distance between a probe dot pair presented in the adapted region and compare them to a reference dot pair presented in a region not affected by adaptation. If the probe dot pair was present inside the adapted area, observers underestimated the distance. If, however, the dot pair straddled the adapted area, the distance was perceived as larger with a stronger distance expansion than compression. Bi-directional spatial repulsion was found with a similar magnitude for size and density adapters. Localization estimates with mouse pointing revealed that adaptation also affected absolute position judgments. Bi-directional spatial repulsion is most likely produced by the lines of adapter stimuli since single bars used as adapters were sufficient to induce spatial repulsion. Spatial repulsion was stronger for stimuli presented in the periphery. This finding explains why distance expansion is stronger than distance compression.

Keywords: aftereffects; space perception; visual adaptation.

Figure 1

(A) Graphical illustration of the trials’ structure in Experiment 1. A trial started with the presentation of a size adapter (8 dva × 8 dva) for 5000 ms. After a blank of 100 ms, the probe and reference stimuli were shown simultaneously for 100 ms, consisting of a dot pair. The probe had a constant inter-dot distance of either 4 dva or 12 dva. (B) The procedure to measure distance judgments after density adaptation. The effects of size and density adaptation were measured in separate sessions. The density adapter consisted of 50 dots that were arranged relative to an invisible grid. Each dot was displaced randomly by 0.3 dva every 300 ms. Except for the features of the adapter stimulus, the procedure was identical as in (A).

Figure 2

Results of Experiment 1. (A) Example psychometric functions from the size adaptation experiment. Data in red show the results from a 4 dva dot pair distance, and data in green results from a 12 dva dot pair distance. (B) Example of psychometric functions from the density adaptation experiment. Same conventions as in 12 dva. (C) Single subject and average distance judgments for the probe dot pairs (4 dva and 12 dva) after size adaptation. Colored symbols show single-subject results and black disks show average results. Error bars are the standard error of the sample mean. The dashed line shows a comparison between the probe and reference pair measured at the baseline. (D) Single subject and average distance judgments for the probe dot pair after density adaptation. Same conventions as in (C).

Figure 3

Results of Experiment 2. (A) Single subject (colored symbols) and average distance judgments (shown in black) for the probe dot pair after adaptation to the upper bar. Error bars are the standard error of the sample mean. The dashed line shows a comparison between the probe and reference pair measured at the baseline. (B) Single subject (colored symbols) and average distance judgments (shown in black) for the probe dot pair after adaptation to the lower bar. The same conventions as in 3A. (C) Single subject (colored symbols) and average distance judgments (shown in black) for the probe dot pair after adaptation to the lower and the upper bar are presented simultaneously. The same conventions as in 3A. (D) Single subject (colored symbols) and average distance judgments (shown in black) for the probe dot pair after adaptation to the lower and the upper bar presented simultaneously at a higher eccentricity. The same conventions as in (A).

Figure 4

Results of Experiment 3. (A) The average localization of probe dots before (black disks) and after adaptation (red disks). Error bars are the standard error of the sample mean. The dashed line indicates the borders of the adapter. (B) Average standard deviations for horizontal localizations of probe positions. The same conventions as in 4A. (C) Average standard deviations for vertical localizations of probe positions. The same conventions as in (A).