1	Perception PART 2 Last lecture: looked at neurophysiology of perception, high-level concepts such as figure-ground and depth cues. Today we’ll finish off high-level concepts go through the lower-level rules that are behind those phenomena I’m not srue exactly waht tihs dmesonaterts but it’s prtety cool. Can you raed tehse setnecens?
2	Perceptual Grouping Wertheimer: Why do some things seem to “go to together?” Some principles of grouping… Will cover grouping further in design lecture
3	Motion Covered in part in Fulvio Domini guest lecture on Optic Flow
4	Roadmap Mystery of vision and historical theories Ecological basis High-level perception concepts Figure-Ground Size constancy Depth and object solidity Lower-level: rules for image “construction” Visual Intelligence [Hoffman 1998] book has 35 rules that the perceptual system uses to decode visual stimulus and create a useful description of the world These underly the higher-level perceptual effects Perception research in the Visual Methodologies [Rose 2001] framework
5	Principles (really conditions) and Rules of Vision Art & design have guidelines–vision has some actual principles and rules Some guidelines confirmed by rules, some not Knowing these can help us Understand how to make the most of the high-level concepts just covered E.g., subtleties of figures in figure/ground organization Better interpret visual materials Design better Understand why some designs fail Explains how much of visual world (and even abstract ideas) can be conveyed with nothing but lines on paper.
6	Principle 1: Images Are Ambiguous “You construct visual worlds from ambiguous images in conformance to visual rules.” [Hoffman 1998 p.24] Image at eye has countless number of possible interpretations. [Hoffman 1998] p13. 2. Even though we have stereo vision, the image at each eye is 2D and has countless interpretations in three dimensions
7	Random Dot Stereograms In a RDS, a repeating pattern is used to make it easier to mistake one element of the picture for another when viewed with crossed eyes The pattern is slightly altered and your (crossed) eyes think they have converged on the same element—when in fact it is two different ones. Effect can happen with repeating wallpapers sometimes
8	Principle 2: Generic (Non-Accidental) Views Construct only those visual worlds for which the image is a stable (i.e., generic) view If changing view slightly changes image dramatically, then was not a stable (generic) view
9	Rule 1: Straight Lines Stay Straight “Always interpret a straight line in an image as straight line in 3D.” [Hoffman 1998 p27]
10	Rule 2: Connections Stay Connected “If the tips of two lines coincide in an image, then always interpret them as coinciding in 3D.” [Hoffman 1998 p27]
11	Lack of Connection…
12	Non-Accidental Line Configurations Many of the rules in Hoffman apply to specific lines configurations in a 2D image These arrangements are statistically highly correlated with non-accidental views of 3D reality
13	Rule 3: Collinear Lines Stay Collinear “Always interpret lines collinear in an image as collinear in 3D.” [Hoffman 1998 p31.]
14	Rule 4: Nearby Elements in 2D are Nearby in 3D “Interpret elements nearby in an image as nearby in 3D.” [Hoffman 1998 p32] Aka a “proximity” effect “As you observe the cube, note that two of the circles seem closer to you, and two further away. When the cube reverses in depth, so do the circles, suggesting that human vision assigns depth to the circles based on the depth it assigns to nearby portions of the cube.” [Hoffman 1998 p32.] This demonstration was devised by Marc Albert, motivated by work of Allan Jepson and Whitman Richards.
15	Rule 5: Smooth curves in 2D are smooth curves in 3D & Rule 6: Curve in 2D = rim of a 3D surface (silhouettes) “Always interpret a curve that is smooth in an image as smooth in 3D.” [Hoffman 1998 p33.] “Where possible, interpret a curve in an image as the rim [contour/edge] of a surface in 3D.” [Hoffman 1998 .p.39] i.e., as bounding part of a surface, not just a curved line hanging out in space
16	Principle 3: 3D World Gets Projected to 2D How 3 dimensions get flattened into 2 on retina—reconstructed to 3D in brain Natural perspective: why moon can block sun Formal rules of linear perspective: Brunelleschi et al (more on this later in course)
17	Rule 7: T-Junction = Rim Concealing Itself “Where possible, interpret a T-junction (aka or J-junction) in an image as a point where the full rim conceals itself: the cap conceals the stem.” [Hoffman 1998 p.39] Explains interposition (overlapping) depth cue from earlier
18	Fun with T-Junctions
19	Three Types of Surfaces Convex Like an egg shell Curves down for observer at any point on surface Concave Like inside of egg shell Curves up for observer at any point on surface
20	Surface Characterizations Surfaces can only curve in these three ways! Convex regions usually bound a material/object Concave regions usually bound pockets of air (aka negative space) Saddle regions bound nothing—instead provide transition between convex and concave (this type of surface often not acknowledged in art theory)
21	Rule 8: Convex bound = convex rim & Rule 9: Concave bound = saddle Rule 10: Smooth it out Rule 8: “Interpret convexities in a silhouette as convexities in 3D.” [Hoffman 1998 p.42] Rule 9: “Interpret concavities in a silhouette as saddles in 3D.” [Hoffman 1998 p.42] Rule 10: “Construct surfaces in 3D that are as smooth as possible” [Hoffman 1998 P. 43]
22	Suggesting 3D Forms with 2D Curves
23	Using the Rules Together d
24	In-Class Exercise—Automatic Birds From [Massironi 2001 p.217] The Psychology of Graphic Images: Seeing, Drawing, Communicating,
25	Constructing vs. “Seeing” (1/2) Construction of objects preconscious—we don’t notice it when it’s working Mr. S [Hoffman 1998 p.49] could “see” but not construct objects Saw color, motion, edges but could not determine what objects they represented Still knew what objects were (could touch, hear them) but couldn’t put the pieces together
26	Constructing vs. “Seeing” (2/2) Construction of objects preconscious—we don’t notice it when it’s working Mrs. B constructed objects that weren’t there [Hoffman 1998 P 76] After stroke could still read, recognize faces but… Vivid, realistic hallucinations of traffic, children, animals, and more
27	Rule 11: Convex Cusps = Occlusion “Construct subjective figures that occlude only if there are convex cusps.” [Hoffman 1998 p.57]
28	Rule 12: Non-accidental Relations = An Object “If two visual structures have a non-accidental relationship, group them and assign them to a common origin.” [Hoffman 1998 p.60]
29	Rule 19 + 20: Figures (and their Borders) have more “salient parts” (i.e., protuberances) than backgrounds. Rules 17 and 18: salience of a cusp increases with angle at cusp and salience of a smooth boundary increases with magnitude of curvature I.e., the more spiky and protruding, the more likely to be “figure” vs. “ground” “Choose figure and ground so that figure has “more salient parts” (rule 20, [Hoffman 1998 p102])
30	Seeing in 3D—Stereo Vision Binocular vision helps give us sense of depth Only noticed recently (relatively speaking) in 1838 by Charles Wheatstone! More in CS graphics lecture
31	Roadmap Mystery of vision and historical theories Ecological basis High-level perception concepts Figure-ground Frame of reference Depth cues Lower-level: rules for image “construction” Visual Intelligence book has 35 that the eye-brain use to decode visual stimulus and create a useful description of the world These are behind the higher level perceptual effects Perception research in the Methodologies framework
32	Perception, A Visual Research Methodology? Approach Take images seriously (goal of Power of Images lectures) yes Consider the social context in which the image was made and is viewed not really Consider your own “ways of seeing” yes (vision mediated by perceptual system, no guarantee we all see the same things) Examine one or more of the three steps in visual message process Production of image (how it is made) ? Not sure Image itself (what it looks like) yes Audience (how it is seen) perhaps not in intended sense (brain injured an audience) Key factors to consider Technological (tools used) ? Compositional (formal design strategies) yes (by nature vs. people) Social (economic, political, social relations, institutions and practices, cultural settings, etc.) somewhat-perceptual findings often relevant
33	Perception Theory Meets Culture Theory in The Science of Beauty How do higher-level issues fit into ecological theories of vision? Science of beauty interesting example: combines issues of Power of images Power of media Gender roles Psychology Evolutionary theory/ecology Gestalt: symmetry, grouping (beauty as symmetry study) Ectoff describes research showing perception of female beauty may be in part hard-wired
34	Facial Symmetry—Better? Almirantis, Y. 1995. Left-right asymmetry in vertebrates. BioEssays Cancar, D. 1995. Sex and the symmetrical body. New Scientist Enquist, M., A. Arak. 1994. Symmetry, beauty and evolution. Nature Etcoff, N. 1999. The Beauty of Science, Survival of the Prettiest Eugene, A. 1998. I Want To Be Beautiful, interview Gould, S. J. 1998. The allure of equal halves. The Sciences Grammer, K., R. Thornhill. 1994. Human (Homo sapiens) facial attractiveness and sexual selection: the role of symmetry and averageness. Journal of Comparative Psychology Symmetry and beauty Mealey, L., R. Bridgstock, G. C. Townsend. 1999. Symmetry and perceived facial attractiveness: a monozygotic co-twin comparison. Journal of Personality and Social Psychology Samuels, C. A., G. Butterworth, T. Roberts, L. Graupner, G. Hole. 1994. Facial aesthetics: babies prefer attractiveness to symmetry. Perception Scutt, D., J. T. Manning. 1996. Symmetry and ovulation in women. Human Reproduction Swaddle, J. P., I. C. Cuthill. 1995. Asymmetry and human facial attractiveness: symmetry may not always be beautiful. Proceedings of the Royal Society of London, Series B: Biological Sciences UTMB, Galveston, TX - Dept of Otolaryngology - Facial Analysis; October 1, 1997 http://www.uni-regensburg.de/Fakultaeten/phil_Fak_II/Psychologie/Psy_II/beautycheck/english/index.htm