Session 1

Think global, act local: Do local visual processing biases explain proficiency in observational drawing in non-autistic individuals?

Rebecca Chamberlain, Chris McManus, Howard Riley, Qona Rankin and Nicola Brunswick

Exceptional graphical abilities in autistic savants have been explained by enhanced local visual processing, coupled with an intact global advantage effect under voluntary selective attention (Plaisted, Swettenham & Rees, 1999, Journal of Child Psychology & Psychiatry, 40, 733-742). Furthermore, it has been suggested that non-autistic children who are precocious at drawing exhibit the same local processing hallmarks as their autistic savant peers (Drake, Redash, Coleman, Haimson & Winner, 2010, Journal of Autism and Developmental Disorders, 40, 762-773). Similar effects have been seen in artistic adults whose drawing experience was correlated with reduced holistic processing in face perception tasks (Zhou, Cheng, Zang & Wong, 2011). In an initial study performance on the embedded figures task (EFT), a measure of local visual processing, independently predicted both self-perceived and objectively assessed drawing ability. This finding was examined in a study that probed both local and global visual processing in art students and controls using Navon shape stimuli, the Block Design Task (Shah & Frith, 1993, Journal of Child Psychology & Psychiatry, 34, 1351-1364) and the attention to detail subscale of the Autism Spectrum Quotient (AQ). The results are discussed with reference to perceptual enhancement theories of observational drawing ability.

Depicting visual perception in art and science

Robert Pepperell

This paper will discuss my attempts to depict visual perception in painting and drawing. Depicting visual perception means trying to capture natural scenes as they are actually perceived rather than as they might appear in, say, a photograph or computer generated rendering. As I will show, there are a number of fundamental features of visual experience that conventional imaging technology does not record. The most important of these are the differentiation between central and peripheral vision and the relative indeterminacy of objects in the periphery, deformations of objects in space relative to viewing position, and the presence of the viewer's own body in the field of view. Once these features are accommodated into the depiction, I argue, we arrive at image that is much closer to actual visual experience than images that conform to linear perspective and omit the appearance of the viewer's own body in the periphery of the visual field. The paper will consider the implications of this approach for the scientific study of perception, how it links to some recent neuroscientific research, and how artists and scientists might benefit from further developing the methods outlined here.

The Mazzocchio in Perspective

Kenneth Brecher

The mazzocchio was a part of 15th century Italian headgear. It was also a kind of final exam problem for students of perspective. In this presentation, we review its history; show 3D models made by hand in wood and using stereolithography in plastic; and discuss a previously unreported visual effect. When viewing a wooden checkerboard mazzocchio - either monocularly or binocularly - that we have fabricated consisting of 16 alternating dark and light segments circumscribing its overall toroidal shape, the outer sides of the individual segments appear to bow inward. This concave inward 'mazzocchio effect' appears to be strongest when viewing a model containing 15 - 20 segments. With too few segments the edges appear flat as they indeed are; and with too many segments, the outer edges appear to take on the convex outward shape of a circle. What causes this 'mazzocchio effect' is not clear. The well known Wundt and Hering illusions that are made from straight lines only (but with intersections) also elicit the percept of curvature. However in the case of the checkerboard mazzocchio, there are no intersecting lines. This research is part of Project LITE which is supported by NSF Grant # DUE - 0715975.

A comparison of statistical regularities in images of regular text, calligraphy and aesthetic artworks

Tamara Melmer, Michael Koch, Joachim Denzler and Christoph Redies

The statistical properties of natural scenes and visual artworks share scale-invariant properties in the Fourier domain. In particular, their radially averaged (1d) Fourier power falls off linearly with a slope of about -2 in log-log plots [Redies et al., 2007, Spatial Vision 21(1-2), 137-148]. In the present study, we asked whether this property can also be found in other categories of man-made images, such as regular text and text with artistic claim (calligraphy). Our results show that the 1d Fourier spectrum of regular text can be roughly divided into two segments. With 8 lines of text per image, the slope of the lower-frequency segment (2-40 cycles/image) is about -1 while that of the higher-frequency segment (40-512 cycles/image) is about -3. The ratio of the two slopes is about 0.3-0.4. This ratio increases with rising artistic claim. It is about 0.5-0.6 for calligraphy and reaches values of about 1 (straight line) or higher for aesthetic artworks. Results were similar for examples from three cultures (Western, Arabic and East Asian). In conclusion, our findings imply that images of artworks and artistic writing contain more global structure (low frequencies) relative to fine detail (high frequencies) than images of regular text.

Getting the shape right: drawings focus on proportion in the positive space

Linda Carson, Matthew Millard, Nadine Quehl and James Danckert

Artists who teach drawing share a conviction that accuracy can be improved by paying careful attention to the spaces between objects ('negative space'). In a geometric analysis of 34 drawings of a complex still life, we applied objective measures of local and global drawing accuracy that were consistent with the expertise of the participants. When we compared each participant's accuracy in drawing shapes in the positive space (figure) and negative space (ground) we found that people made similar errors of orientation, position and scale in both. As drawing teachers would predict, though, participants made significantly larger errors of proportionality in the negative than in the positive space. That is, people did much worse at 'getting the shapes right' in the ground than in the figure, despite the fact that there were no figure/ground differences in orienting, positioning or scaling those shapes. Even expert participants, who were more accurate than novices on every dimension of error, made larger errors of proportionality in negative than in positive space. We discuss the implications for drawing education and how these error measures can be used to probe for similar differences in visual perception.

Nearly random naturalistic textures from photographs and paintings

Jacques Ninio

In previous work (Ninio, 2007 Spatial Vision 20 561-577) I showed how to design camouflaging textures having as in real life edges at all orientations. These textures were suitable for covering curved surfaces without visible join. The starting material could be photographs of a collection of similar objects (for instance, barks or sea-shells, or natural heaps of objects such as leaves or seaweed). The starting textures could also be produced manually by mixing, according to various protocols, modeling pastes of various shades of gray. The initial textures were then transformed by cut and paste procedures to generate suitable final textures. When symmetry is introduced in these textures, meaningful scenes emerge vividly, whatever the position of the symmetry axis. The work has now been extended by using new photographic material, and also reproductions of paintings from Tiepolo, Gauguin, Kokoshka, Klimmt, Ligabue and others. Areas in the paintings in which patches of different colors converge were extracted and assembled into abstract textures. Occasionally, these abstract textures provided, even without symmetry, a sense of pictorial realism. It could be due to the presence of a good balance between filled and empty spaces, that was present in the paintings, and absent in the author's initial textures.