From Flashes to Edges to Objects: Recovery of Local Edge Fragments Initiates Spatiotemporal Boundary Formation

Abstract

Spatiotemporal boundary formation (SBF) is the perception of illusory boundaries, global form, and global motion from spatially and temporally sparse transformations of texture elements Kellman, 1993a, 1994;Erlikhman and Kellman, 2015). It has been theorized that the visual system uses positions and times of element transformations to extract local oriented edge fragments, which then connect by known interpolation processes to produce larger contours and shapes in SBF. To test this theory, we created a novel display consisting of a sawtooth arrangement of elements that disappeared and reappeared sequentially. Although apparent motion along the sawtooth would be expected, with appropriate spacing and timing, the resulting percept was of a larger, moving, illusory bar. This display approximates the minimal conditions for visual perception of an oriented edge fragment from spatiotemporal information and confirms that such events may be initiating conditions in SBF. Using converging objective and subjective methods, experiments showed that edge formation in these displays was subject to a temporal integration constraint of ∼80 ms between element disappearances. The experiments provide clear support for models of SBF that begin with extraction of local edge fragments, and they identify minimal conditions required for this process. We conjecture that these results reveal a link between spatiotemporal object perception and basic visual filtering. Motion energy filters have usually been studied with orientation given spatially by luminance contrast. When orientation is not given in static frames, these same motion energy filters serve as spatiotemporal edge filters, yielding local orientation from discrete element transformations over time. As numerous filters of different characteristic orientations and scales may respond to any simple SBF stimulus, we discuss the aperture and ambiguity problems that accompany this conjecture and how they might be resolved by the visual system.

Spatiotemporal and Luminance Contrast Properties of Symmetry Perception

“…That is, elements revealed at different times and locations can be assembled into unitary perceived objects [22,23]. The temporal integration of spatial information appears to be constrained to a temporal integration window between 80 ms [24] and 165 ms [21], which is in keeping with findings showing that temporal integration of symmetry occurs within~120 ms [19].…”

“…This is to say that the perception of such a target requires a spatiotemporal integration process. Dynamic element changes over time and space are sufficient to define boundaries and surfaces [20,21], occluded objects [22], and illusory shapes [23,24]. That is, elements revealed at different times and locations can be assembled into unitary perceived objects [22,23].…”

Bilateral Symmetry Strengthens the Perceptual Salience of Figure against Ground

“…The new effect found here, where symmetry strengthens the figure-ground salience of surfaces from occlusion cues, which form through visual spatial interpolation across fragments within a narrow temporal window of processing [44,45], fully expresses the adaptive logic of visual preference for symmetry. Symmetry enables perception-based decision making, and survival relevant responses to symmetry or lack thereof can be found in animal species [37].…”

“…Variations in symmetry could be tested against variations in the support ratio in a first instance. Figure-ground segregation from interpolation is an early-stage process in perceptual grouping [16,44,45], as is symmetry detection [39,43]. In particular, as shown by Erlikhman and Kellmann [44,45], the human perceptual system uses critical spatial cues of local position and alignment within a restricted spatiotemporal window (~165 msec) for the rapid extraction of co-oriented edge fragments from the visual input.…”

Color for the perceptual organization of the pictorial plane: Victor Vasarely's legacy to Gestalt psychology

“…Specific multi-level interactions between detectors accomplish such hierarchical resolution of uncertainty in response to physical input, thereby generating the visual percepts that we consciously see. The general conclusion here is that the brain does not eliminate uncertainty too soon, but takes advantage of uncertainty until processing stages are reached at which uncertainty can profitably and drastically be reduced or eliminated ( Erlikhman and Kellman, 2016 ; Silvestri et al, 2010 ). In the case of figure-ground perception, when all other cues to visual depth in two planar configurations are equal, the difference in temperature of a local color may become a strategic cue to pictorial depth, as successfully captured and quantified in this work here.…”