Perception, Part I
‹date/time›
Click to edit Master text styles
Second level
Third level
Fourth level
Fifth level
‹footer›
‹#›
SO, last week we looked at cultural theory and linguistically approaches to decoding images. These theories (semiotics, discourse theory, general cultural theory) are valuable tools for understanding the role of images in our culture, how we may be influenced by them, and how to better bring to SEE them and understand them.
As we noted with the “gorilla” exercises, we are not always aware of what we are seeing and usually don’t consciously analyze it.
This week we will learn something about how our brains perform the feat of seeing and where in this process image interpretations of different types occur. ON the one hand, we’ll see that , in a way, it’s ALL interpretation, and that, as the author of VI claims, we actively construct our visual world. On the other hand, we’ll see that this does not mean that reality is subjective or possibly non-existent.
We’ll start with some past and present theories of vision, proceed to some topics you’ve probably heard of, if not studied, such as figure-groudn distinctions, and then dip into the lower-level “rules” described in the VI book.
Finally, we’ll assess this way of looking at images in terms of the Framework proposed by Gillian in the VMs book.
Interesting factoid, from a book by James Elkins called
Visual Studies, a Skeptical Introduction.
This is where the money is and most of the manpower. 
this and the technical CG section is where this course is probably most different from courses you will find on modern culture, culture studies, or visual culture.
First of all, even now, it is not at all obvious how vision works. Much to still be discovered.
SOURCES
___________________
http://fnordfiles.rtfm.com.au/graphics/Art/Escher/escher.htm
http://bulldog2.redlands.edu/fac/jeremy_anderson/research/Epicurus.pdf Epicurus (1964). Letter to Herodotus, in Epicurus: Letters, Principal Doctrines, and Vatican
Sayings, Russel M. Geer, trans., New York: Macmillan.
Or
Epicurus (1926). Epistula ad Herodotum, in Epicurus: the Extant Remains, Cyril Bailey, trans.,
Oxford: Clarendon Press.
Lucretius (1975). De Rerum Natura, W. H. D. Rouse,
Cambridge: Harvard University Press.
Surfaces oriented (bring in Tinker Toy prop or other prop)
Bring in bowl
Bring in more knowledge of real world as you go up this hierarchy
http://corvinae.blogs.sapo.pt/arquivo/242781.html
SOURCES
_____________________
[Palmer 1999]
http://corvinae.blogs.sapo.pt/arquivo/242781.html (bird photo)
Gestalt theory by Max Wertheimer, http://gestalttheory.net/archive/wert1.html#fn1
(“Isolate the elements, discover their laws, then reassemble them, and the problem is solved.”) Max Wertheimer, 1924 http://gestalttheory.net/archive/wert1.html#fn1
http://www.yorku.ca/eye/brain2.htm
“Visual images are inverted as they pass through the lens. Therefore, in your right eye, the nasal retina sees the right half of the world, while the temporal retina sees the left half of the world. Notice also that the right nasal retina and the left temporal retina see pretty much the same thing. If you drew a line through the world at your nose, they would see everything to the right of that line. That field of view is called the right hemifield. So, what you see is divided into right and left hemifields. Each eye gets information from both hemifields. For every object that you can see, both eyes are actually seeing it - this is crucial for depth perception - but the image will be falling on one nasal retina and one temporal retina. Why bother to divide the retinas at all? Recall that the brain works on a crossed wires system. The left half of the brain controls the right side of the body, and vice versa. Therefore the left half of the brain is only interested in visual input from the right side of the world. To insure that the brain doesn't get extraneous information, the fibers from the retina sort themselves out to separate right hemifield from left hemifield. Specifically, fibers from the nasal retinas cross over at the optic chiasm - whereas the temporal retinas, already positioned to see the opposite side of the world, do not cross. Here is what it looks like: “http://thalamus.wustl.edu/course/basvis.html Neuroscience tutorial at Wash U at St Louis
To electrochemical signals (true?)
Neural retina from adult human, showing how nuclear laminae alternate with layers of processes. Photoreceptor cells are at the bottom of the figure. Image courtesy Janet Sparrow  http://cpmcnet.columbia.edu/dept/neurobeh/vision-grant/
Eye diagram: The retina is covered with rods and cones, light-receptors that send electrical impulses through optic nerve fibers and to the brain.  http://www.visionweb.com/content/consumers/dev_consumerarticles.jsp?RID=36
Like digicam CCD
Terabytes of data generated by tiny fraction of output of a whole retina. Brain does awesome data processing job.
The goal of the Retina Project is to understand the language the eye uses to send information about the visual world to the brain, to study how a biological neural network (the retina) processes and encodes information. This is a truly interdisciplinary collaborative effort between high energy physicists from SCIPP, and neurobiologists, experts in nanofabrication, and VLSI designers. The retina is a thin tissue which lines the back of the eye. It is a sophisticated pixel detector that converts an input visual image into a set of highly processed electrical signals ("spikes") that travel up the optic nerve to the brain. These spike trains collectively encode and communicate the critical features of the image in a timely manner to the visual cortex. To study this neural code, we are developing an imaging system for large-scale neural activity. This "Retinal Readout System" will allow the simultaneous recording of signals from hundreds of retinal output neurons (the ganglion cells) as a dynamic visual image is focussed on the input neurons (the photoreceptors). The implementation of this system is based on the silicon microstrip detector technology and expertise we have pioneered and developed in SCIPP to study short-lived particles and CP violation, and to search for the Higgs Boson, in high energy physics experiments. In the neurobiology experiment, live retinal tissue is placed in a chamber on top of an array of microscopic electrodes fabricated on a glass substrate. The chamber is filled with physiological saline solution; retinal tissue superfused in this manner can be kept alive for several hours. A dynamic image on a computer display is focussed on the photoreceptors. In response to this image, the ganglion cells generate spikes which are picked up as extracellular signals by the electrodes. These signals are amplified, filtered, digitized and recorded. The spike trains from the hundreds of neurons detected can then be correlated in space and time among themselves and with the visual image to study the neural code at the retina/brain interface.
http://www.snl-e.salk.edu/technology/
“retinal map” is highly nonlinear, further processing must undo distortion to determine real-world stimulus
This clearly shows parallel processing and some degree of locality
Role of these bizarre stories: to show how “vision” is many processes in the brain, with whole objects/motion arising from separating processing of form, color, motion, parts, and wholes…With advanced medical imaging and other techniques, vision science has taken off in the last 25 years and much new research has been done.
Still a great deal that we do not know, however.
Connection to other materials here.
Simplified in shape, color, texture, etc.: no “perfect (mathematical) sphere” in real world…
James J. Gibson, The Perception of the Visual World (1950). The Ecological Approach to Visual Perception (1979),
Rather than study what confuses the eye-brain, Gibson focused on the amazing fact that in the complex visual work we live in, we are rarely mistaken about what we see. What rules govern our ability to make sense of this massive visual input??
For us, how can we take advantage of this powerful system we’ve evolved to help us communicate, think better?
Constantly checking visual data against other senses and “reality” checks to influence what we see and how we feel about it.
Also seen in phenomenon of “acclimation” -- something may look strange or alarming, but if it turns out not to be, we quickly come to ignore it.
Helmholtz's Treatise on Physiological Optics is widely recognized as the greatest book ever written on vision.
Helmholtz links
http://elvers.stjoe.udayton.edu/history/people/Helmholtz.html
Tree: ground is continuous,, tree is continuous
Tree straight line, right angle to ground
Parallel sides of trunk indicate parallel side in real world
Cheetah-symmetric, lighting from above,
Lots of dots but we see as one coherent animal. Grass has texture but we can discern plane it is on.
Left: titled target “pops out” from field of straight lines.
Opposite effect (right) is not so pronounced.
Avd and “leaning” effect
Forward ref to “light comes above “ rule
For ref laidlaw lecture
VS. p555
As we noted before in the physiology section…
Fix progression
OUTLINE SALMON oON BOTTOM
Portions of figure 12 and 13, from Healey, C. G., Booth, K. S., and Enns, J. T. "Visualizing Real-Time Multivariate Data Using Preattentive Processing." ACM Transactions on Modeling and Computer Simulation 5, 3, (1995), 190-221. http://www.csc.ncsu.edu/faculty/healey/download/tomacs.95.pdf
During this week , we’ll be conscious applying some of these perceptual ideas to understand the process of drawing and graphical communication. We’ll also try to integrate the visual studies/semiotics theories with what is known about the science of vision.
http://www.michaelbach.de/ot/
http://www.owlnet.rice.edu/~psyc351/Images/IllusoryRotation2.gif
Lateral inhibition? After images?
EXPLANATION??
See syllabus/Web links.
Andy to talk about Escher
SOURCES
_________________
http://psylux.psych.tu-dresden.de/i1/kaw/diverses%20Material/www.illusionworks.com/html/impossible_trident.html
http://catb.org/~esr/jargon/html/B/blivet.html
http://neuro.med.harvard.edu/site/dh/b19.htm
Attention is also how we combine the features we discussed in pop out section into an object –and unite “line” with “diagonal” with “color”, etc. .
Also, the more you look, the more you see… ref visual memory exercise earlier. Can bring preconscious processing into consciousness…
Need to distinguish object (hungry tiger) from background (harmless grass) very quickly.
We’ll be looking at some lower-level reason for calling one thing a figure and another ground in the next section of the talk.
Explain optic flow briefly
SUPPLIES NEEDED:
Handout with face invariants drawn
Pencils
Eraser
Blank paper
Not 100% definitie—add grid to see.
ADD GRID
Which circle looks like disk? Sphere?
Grinch-like figure from http://www.cs.stevens.edu/~quynh/courses/cs638-sp05/drseuss_kowalski.pdf
PROGESSIVIZE
Vision: extracting shape from shading
Artists use this in induce 3d—literally creatign sense of shape from shading
http://psylux.psych.tu-dresden.de/i1/kaw/diverses
%20Material/www.illusionworks.com
/html/size_constancy.html
This site often down, but good stuff when it’s up