Spatial frequency integration during active perception: perceptual hysteresis when an object recedes

Abstract

As we move through the world, information about objects moves to different spatial frequencies. How the visual system successfully integrates information across these changes to form a coherent percept is thus an important open question. Here we investigate such integration using hybrid faces, which contain different images in low and high spatial frequencies. Observers judged how similar a hybrid was to each of its component images while walking toward or away from it or having the stimulus moved toward or away from them. We find that when the stimulus is approaching, observers act as if they are integrating across spatial frequency separately at each moment. However, when the stimulus is receding, observers show a perceptual hysteresis effect, holding on to details that are imperceptible in a static stimulus condition. Thus, observers appear to make optimal inferences by sticking with their previous interpretation when losing information but constantly reinterpreting their input when gaining new information.

Keywords: hybrid images; hysteresis; perceptual organization; spatial frequency.

Figure 1

Example display. The hybrid image in the middle is composed of the left image (in the low spatial frequencies) and the right image (in the high spatial frequencies). Observers had to stand at the distance indicated and judge whether the hybrid looked more like the left or right component image using a keypad. Because of the human contrast sensitivity function (see Appendix), when viewing this hybrid from close up it should look like the right image; when holding it far away or squinting it should look like the left component image.

Figure 2

Data from Experiments 1 and 2. The red lines indicate the percentage of trials on which observers in the static condition reported seeing the high frequency image (error bars are ±1 SEM). The green triangles and blue squares show the distance/visual angle at which observers reported the point of subjective equality in the dynamic conditions (error bars are ±1 SEM). When starting far from the image observers’ reported lined up perfectly with their reported point of subjective equality from the static conditions. When starting close to the image, observers tended to stick with the high frequency image considerably longer than predicted by their static responses.

Figure 3

Example display from Experiment 3. The images on the left and right of the display are the images composing the hybrid image located in the center of the display. Observers’ task was to stand at the distance indicated and report whether the hybrid looked more like the image on the left or the image on the right.

Figure A1

(A) The information contained in a hybrid image is a mixture of the information originating from the high spatial frequency component of an image (blue) and the low spatial frequency component of another image (green). (B) From a distance of 3 m, the information visible to human observers (the information available from a standard CSF, depicted as the dashed-gray line) is predominantly driven by the low spatial frequency component: observers should preferentially perceive the happy face. (C) From a distance of 0.5 m, the visible information is predominately from the high spatial frequency component: as a result, observers should see a neutral face.

Figure A2

(A) The total amount of information available from each component after taking a weighted sum by the CSF at a given distance. The y-axis is normalized to put the maximum information at 1.0. (B) The percentage of all information about the image available at a given distance that comes from the high spatial frequency component.