The Psychophysics of Appearance Laboratory

University of Bern, Institute of Psychology

What are the mechanisms that generate our rich and stable visual world from the highly ambiguous and ever-changing light pattern falling on the retina? How do complex interactions between local and global levels of processing, and their dependence on factors such as grouping and Gestalt, shape visual perception? Why do we experience the visual world the way we do?

In the Psychophysics of Appearance Laboratory, we investigate these and related questions, primarily in the realm of spatial vision. For example, to understand the role of perceptual organization and context in visual perception, we study crowding, masking, and visual search using a variety of psychophysical procedures. We extend traditional performance measures by directly quantifying perceptual experience using novel appearance-based methods, such as analyzing observer drawings of gaze-contingent peripheral stimuli. Finally, we foster connections between scientific and artistic explorations of the visual world, demonstrating how vision science and art can mutually enrich each other.

Contextual modulation and crowding

The perception of a target stimulus is strongly modulated by its context. One example, "crowding," refers to the phenomenon whereby objects are more difficult to perceive when surrounded by other objects, a condition especially pronounced in the visual periphery. Several of our studies show the powerful influence of perceptual organization, challenging current explanations of crowding.

Crowding and visual search

(with Johan Wagemans)
Whereas most typical crowding experiments require focal attention to a known location in the visual field, a more ecologically valid task may be one in which object locations are unknown. Here we tested the ability to determine the polar location of an oddball letter in a ring of crowded peripheral letters.

Haptic perceptual organization

(with Krista Overvliet)
Since the early phenomenological demonstrations of Gestalt principles, one of the major challenges of Gestalt psychology has been to quantify these principles. Here, we show that contextual modulation, i.e. the influence of context on target perception, can be used as a tool to quantify perceptual grouping in the haptic domain, similar to the visual domain. We conclude that in haptics, similar organization principles apply as in visual perception and argue that grouping and Gestalt are key organization principles not only of vision, but of the perceptual system in general.

Crowding and MOT

(with Johan Wagemans and Lee de-Wit)
In dynamic, cluttered environments, humans often track several moving objects simultaneously. What are the limits of this multiple object tracking (MOT) task? It has been proposed that tracking is limited by the number of objects, their speed, or their spacing. Here, we investigated the role of crowding in MOT and found that crowding-specific spatial relations – beyond spacing per se – is one of the factors limiting MOT.

Foveal repetitions uncrowd

(with Patrick Cavanagh and John Greenwood)
An item presented in the fovea that matched a peripheral, crowded target improved target discrimination compared to a non-matching item, indicating that crowded target signals are not irretrievably lost but can be made visible by long-range, shape-specific grouping.

Backward masking and contextual modulation

(with Michael Herzog and Mauro Manassi)
The strength of visual backward masking depends on the stimulus onset asynchrony (SOA) between target and mask, as well as the spatial layout of target and mask. Here, we investigated in how far backward masking is similarly affected by grouping processes as crowding. Similar to crowding, we show that masking is reduced when the flankers are part of a larger spatial structure. Complex Gestalts seem to need longer processing times to show ungrouping effects as observed in crowding.

Grouping versus crowding

(with Patrick Cavanagh)
A number of studies have shown that crowding shares many features with grouping. Here, we investigated whether effects of crowding and grouping on target perception are related by asking whether they operate over the same spatial scale. We found that grouping operates over larger spatial scales than crowding, indicating that crowding and grouping are different processes. We suggested that target-flanker grouping changed the target’s appearance by shape specific assimilation.

Conspicuity, crowding, and visual search

(with Michael Herzog and Gerald Westheimer)
Target-flanker grouping plays a major role in crowding. Here, we investigated the relationship between target-flanker grouping (or “standing out” of the target) and performance using two independent tasks on the same stimulus configurations. Our results show that targets uncrowd (as measured with vernier offset discrimination) when they pop out (in visual search).

Gestalt factors in basic spatial vision

(with Michael Herzog and Gerald Westheimer)
Since grouping of pattern elements into configurations (Gestalten) was first proposed by Wertheimer in 1923, the ongoing search for the rules governing such grouping processes has been only moderately successful, in part because of the lack of appropriate research tools. Here, we show that Gestalt factors play an important role in contextual modulation of basic spatial vision, and propose that vernier-offset discrimination with different flanking configurations can be used as a quantitative measure to investigate the rules governing the grouping of elements into meaningful wholes.

Malleability of attentional capture

(with Michael Herzog, Anna Grubert, and Joseph Krummenacher)
Onset distractors are thought to strongly elicit bottom-up attention and to capture eye movements. Varying the display probabilities of the onset distractors strongly modulated attentional capture. We suggest that modulation was due to statistical learning. This study adds further evidence that bottom-up processes are not completely automatic.

Contextual modulation and grouping

(with Michael Herzog and Gerald Westheimer)
Here, we showed that grouping (and ungrouping) of target and flankers by contrast polarity, color, and stereoscopic depth determine performance in a foveal vernier discrimination task. The results show that local interactions between the target and the flankers do not explain contextual modulation, but that grouping processes are key. Recently, we found very similar results in peripheral vision (crowding).

Appearance-based psychophysical methods

Kurt Koffka famously asked “why do things look the way they do?” In the Psychophysics of Appearance Laboratory, we investigate a related but logically prior question, namely, “how do things look”, especially when recognition is difficult or ambiguous. We use state-of-the-art technology (e.g., eye trackers for gaze-contingent displays, digital drawing pens and tablets) to study quantitative measures of appearance, and utilize computational methods to analyze and model behavioral results.

Rey-Osterrieth complex figure drawing in the normal periphery

(with Johan Wagemans)
We studied the drawings of art students depicting the Rey-Osterrieth Complex Figure, which was presented peripherally at 12 and 6 degrees. Error patterns had several characteristics of crowding, including number errors and target diminishment.

Identity Crowding

(with Henry Taylor)
It has been proposed that "identity crowding" (crowding when the target and flankers are identical) is an example of seeing without attention. Using unconstrained appearance capture, we showed that performance in identity crowding is not superior to “normal “ crowding. We suggest that cognitive inferences and biases based on prior knowledge as well as perceived regularity of the stimulus explain seemingly better performance in identity crowding.

Characterizing peripheral errors

(with Johan Wagemans)
Using gaze-contingent stimulus presentation in the periphery and an unconstrained drawing method, we characterized the “errors” occurring in crowding. Frequent omissions, truncations, and distortions of elements indicate that crowding involves a broad spectrum of “perceptual errors” that are not revealed in standard crowding paradigms. We proposed that gaze-contingent drawings is a useful tool to investigate crowding.

Art and Vision

We investigate topics at the intersection of visual arts and vision science, ranging from collaborations that enable us to answer scientific questions with the help of artistic skills, to conjoined explorations of contemporary art practices and to using artworks as stimulus spaces that inform us about the functions of the visual system.

Transparency

(with Patrick Cavanagh)
Artists throughout the ages have discovered a number of techniques to depict transparency which closely follow the properties of physical transparency. Here, we show how artists employed these and other techniques revealing their anticipation of current formalizations of perceived transparency, and we suggest new, as-yet-untested principles.

Peripheral Painting

(with Tilde van Uytven and Erik Myin)
Many artists and scientists share an interest in peripheral vision. While vision scientists often investigate peripheral vision by measuring performance on simple stimuli under highly controlled conditions, artists are usually concerned with the appearance of complex objects and scenes when viewed outside the center of gaze. Here, we investigate drawings and paintings of the visual periphery from an artistic and a scientific point of view, showing how peripheral vision and crowding change appearance.

Line drawings

(with Patrick Cavanagh)
Line drawings often yield vivid three-dimensional impressions of scenes and objects even though they only use a sparse set of contours. Here, we ask why the visual system processes line drawings the way it does, and show how line drawings exploit the underlying neural codes of vision. We suggest that line drawings show how artists’ intuitions go well beyond the understanding of vision found in current neurosciences, psychology, and computer vision.

Art projects to investigate phenomenal reality

(with Ivana Franke )
An unusual and misleading lack of correspondence between a perceptual object and an object in the external world is often referred to as 'perceptual error', or as 'non-veridical' perception. With a range of installations characterized by a reduced availability of potentially corresponding external world categories we challenge these concepts. We suggest that the installations are useful for experiential and theoretical education, investigating vision and phenomenal reality.

Training non-conceptual depiction using peripheral vision

(with Tilde van Uytven and Erik Myin)
An important aspect of learning to draw or paint is counteracting the powerful force of conceptual knowledge. We trained a cohort of art students to more faithfully depict their unbiased percepts by having them copy paintings presented peripherally.

Prior knowledge modulates peripheral color appearance

(with Tilde van Uytven and Erik Myin)
Color perception deteriorates with increasing eccentricity in the visual field. Here, we investigated peripheral color perception using a painting method, asking how prior knowledge affects color appearance under crowding conditions. Our results indicate that prior knowledge of peripheral targets strongly shapes perception.