Visual Pattern Perception
Demonstrations (Direct Links)
Demonstration 5.1 Identification of Features in Simple and Complex Figures
Demonstration 5.2 Gestalt Organizing Principles
Demonstration 5.3 Figure-Ground Relationships
Demonstration 5.4 Is an Averaged Face Attractive?
Before You Start
• As you look around you, notice that you don't really see edges and lightness/color differences. Instead, you see books, chairs, people. How does the information that we extract from our environment lead to the perception of such objects? That's the key question that motivates many researchers in visual perception. If you're like most people, your experience is that perceiving objects is immediate and effortless. Nonetheless, the processes that underlie your perceptual experience are many and complex.
• In order for us to survive in this world, we must perceive it fairly accurately. However, it's also the case that our visual system can misperceive an object in the world, or perceive the same stimulus in multiple ways. You should find these experiences intriguing. They are also potentially quite informative about the way our visual system perceives objects.
• Faces are among the more important objects we encounter. Because of their importance, we're quite good at perceiving faces. However, it's also the case that people are notoriously poor at eyewitness identification. How might we resolve the discrepancy between our face perception abilities and the difficulties we face when attempting to remember and identify a face that we may have seen only briefly?
Spatial Frequency Analysis Approach
Link - An intuitive explanation of Fourier theory (Lehar)
Inspect the two diagrams below. In each diagram one figure is deviant because it is different from the other three figures. Do you identify the deviant figure faster in the diagram on the left or the diagram on the right? What might that difference tell you about how we process shape stimuli?
Our experience of the world is quite organized. That is, in spite of the fact that our visual experience is based on the building blocks of receptive fields of varying complexity, we don't experience receptive fields. Instead we perceive objects such as people, desks, etc. What happens to the visual input between receptive fields and objects? The answer to that seemingly simple question is quite complex. One piece of the answer may lie in the Gestalt organzational principles.
As your textbook indicates, people have identified quite a few Gestalt principles of organization. Below you will see figures that illustrate a number of these principles.
(We organize together pieces that move together. Thus, we see the red dots as a unit moving together. However, we see three of the green dots as a unit distinct from the other three green dots, because the three dots move together.)
According to uniform connectedness, we organize as a single unit those parts of the array that appear to be connected. How do you see each of the four examples? Can you see the operation of uniform connectedness? On your lab worksheet, note how uniform connectedness might conflict with other Gestalt principles in the various examples. How does your ultimate experience of the stimuli help you to understand the functioning of the various Gestalt principles? [You can learn more about the principle of uniform connectedness by reading: Palmer, S. & Rock, I. (1994). Rethinking perceptual organization: The role of uniform connectedness. Psychonomic Bulletin & Review, 1, 29-55.]
How do you organize the following nine stimuli? Do you see the squares as rows, columns, or neither? Which Gestalt principles seem to be operating to lead to that organization? Note that some of the stimuli pit one principle (e.g., proximity) against another principle (e.g., similarity).
1. 2. 3. 4. 5. 6. 7. 8. 9.
Link - The face/vase illusion is fairly ubiquitous. What is less common, however, is the opportunity to have your face the one in the illusion. If you'd like to acquire some art that puts you into the face/vase illusion, check out this site.
Demonstration 5.3 Figure-Ground Relationships
Which is figure and which is ground in this picture? Probably, the cross with the radial marks looks like the figure, and the one with the concentric circles looks like the ground, continuing behind the figure. With some effort you can force the radially marked cross to become the ground and the concentrically marked cross to become the figure. In some cases, such as the Rubin face-vase illusion, it's relatively simple to change figure to ground and vice versa. Why is it more difficult to do so in this case? It's probably due to the fact that it's easier to organize the cross with the concentric circles as ground because the circles can be perceived as going behind the radially marked cross. When you perceive the cross with the concentric circles as figure, it's much more difficult to imagine the radial lines going behind the cross.
Link - Christopher Healy (North Carolina State University) has provided a page that illustrates many of the principles of the Feature-Integration approach.
Perceiving Letters in the Context of Words
Perceiving Objects in the Context of Scenes
Distortions of Shape Due to Context
The Role of Time in Shape Perception
• One classic ambiguous (multistable) figure is the Young Girl-Old Woman illusion. Boring (1930) reported a variant of this illusion in the psychological literature, but an earlier variant had appeared as early as 1890.
• M. C. Escher is probably the best-known producer of impossible figures. You can find many examples of his work on the web:
The Official M. C. Escher site
Zvi Har'El's M.C. Escher Collection
David McAllister's Escher Collection (see David's own art).
World of Escher
Andrew Lipson's LEGO site, wherein he shows five Escher constructions with LEGOs. Cool beans!
Link - D. L. "Rusty" Rust is another artist who creates images that are multistable or illusory.
Effects of Unusual Stimulus Orientations
• We can recognize an object even when we see it from different perspectives. However, some perspectives are more challenging than others. The bottle is more difficult to identify when seen from directly above. Even though some of the other viewpoints are atypical, they lead to easier identification.
Photo by H. Foley
Researchers argue about what makes a face attractive. Some of them believe that a face that is an average of a number of individual faces is more attractive than any of the individual faces. Others believe that there is more to an attractive face than being a composite average of a number of indivudal faces. The researchers at The Face Research Lab have been generous enough to provide a web page that allows you to create a face that is the average of a number of individual faces. Try out their web page (Make an Average Face link) to determine if the averaged faces seem more attractive to you. What factors do you think would make a face more attractive?
Link - The Face Recognition Homepage has useful links to databases of images of faces, etc., as does the Face Detection Homepage (which also has a link to an older version of the Face Recognition Homepage).
Link - If you'd like to see faces as they age, check out this site that shows Diego Golberg and his family from 1976 to date.
Link - If you'd like to see what an "average" face looks like, you can create your own average face. Do you think that your average face is attractive? At this site, you can also play with faces in other ways at this site.
Link - Beautycheck is a site that illustrates facial attractiveness via averaging.
Link - Tony Little (University of Stirling) maintains a page that reflects his interests in face perception.
Link - The University of California @ Santa Cruz's Perceptual Science Lab Facial Analysis page has lots of useful links and demonstrations (e.g., McGurk effect).
Link - EvoFit is a face composite system that is under development.
1. Describe bottom-up (data-driven) processing as if you were talking to a student studying introductory psychology. Use your own words in this description, and use examples from the material in this chapter. Which approaches to shape perception seem most dependent on data-driven processes? What role do these processes play in the particular approaches?
2. Throughout this book, we emphasize the importance of top-down (conceptually driven) processing. What is top-down processing? Which approaches to shape perception seem most dependent on these processes?
3. In Chapter 3 we examined the anatomical and physiological bases for visual perception—essentially the hardware of vision. Evaluate each of the approaches to shape perception in terms of the extent to which it is consistent with our knowledge of the anatomy and physiology of vision.
4. Theme 2 of this text emphasizes the importance of the rich context within which we perceive objects. Use several examples from this chapter to support the importance of context in shape perception.
5. You are driving along a country road. Next to a farmhouse you see a crude hand-lettered sign, “EGGS FOR SALE.” Describe how the following approaches account for your recognition of letters in that sign: the spatial frequency analysis approach, template-matching approach, prototype-matching approach, computational approach, and feature-integration approach. Do you think you would find it more difficult to read the sign if you saw it in front of an urban apartment?
6. Think of yourself as you are reading this book. How does the word-superiority effect facilitate reading? How does context facilitate reading? How are saccades involved in reading?
7. Draw an example of a figure with a subjective contour. How would you explain the subjective contour in your drawing? What problems do subjective contours pose for the various approaches to shape perception?
8. Why are ambiguous (multistable) images so important for Gestalt psychologists? Why are these images problematic for approaches to shape perception? In your daily life, you likely see very few shapes that are ambiguous. Why are most shapes not ambiguous?
9. We’re quite good at perceiving faces, and we can recognize hundreds of faces. At the same time, as we discussed in Chapter 2, people are not particularly reliable eyewitnesses. How might you reconcile our prowess at face perception and recognition with our inability to identify perpetrators of crimes?
10. Given all that you’ve learned up to this point, provide your own description of the processes involved in shape perception. Start with a distal stimulus in an unusual orientation, such as a person swinging upside down on some monkey bars. Because of the swinging, that person’s face will not fall on one spot on your retina. What approaches do you think would be most useful to you in explaining how you come to recognize the person?
Thomson Higher Education has published two very useful CD-ROMs. John Baro (Polyhedron Learning Media) has developed Insight: A Media Lab in Experimental Psychology [see Form and Motion] and Colin Ryan (James Cook University) has developed Exploring Perception [see Module 2].
Link - Art, Vision & The Disordered Eye is a useful resource for exploring the intersection of art and perception (in this case, abnormal perception).
Link - V. S. Ramachandran (UC San Diego) has provided a number of illustrations of principles giving rise to shape (e.g., shading, motion).
Link - If you are interested in making use of images (especially of faces) for classroom demonstrations, you'll appreciate knowing about the page of links called Computer Vision Test Images, even though some links are broken. These images are found with other material at the Computer Vision Homepage.
Link - The Vision Science site has become a compendium of links to vision researchers and related materials (another source of images).
Link - Another source of information about vision (especially computational vision) and images is CSAIL (Computer Science and Artificial Intelligence Laboratory) at MIT. (See especially the CBCL images of faces, cars, etc.)
Link - The Pattern Recognition Files are no longer being maintained and may soon disappear, but you may find useful information therein.
Link - Josh McDermott and Ted Adelson (MIT) have produced a very nice tutorial page regarding motion and form perception.
Bruce, V. & Young, A. (1998). In the eye of the beholder: The science of face perception. New York: Oxford.
Frisby, J. P. & Stone, J. V. (2010). Seeing: The computational approach to biological vision (2nd Ed.). Cambridge, MA: MIT Press.
Hoffman, D. D. (1998). Visual intelligence: How we create what we see. New York: Norton.
Jacob, P. & Jeannerod, M. (2003). Ways of seeing: The scope and limits of visual cognition. New York: Oxford.
Jenkin, M. R. M. & Harris, L. R. (Eds.). (2006). Seeing spatial form. New York: Oxford.
Peterson, M. A., Gillam, B., & Sedgwick, H. A. (Eds.). (2007). In the mind’s eye: Julian Hochberg on the perception of pictures, films and the world. New York: Oxford.
Peterson, M. A. & Rhodes, G. (Eds.). (2003). Perception of faces, objects, and scenes: Analytic and holistic processes. New York: Oxford.
Regan, D. (2000). Human perception of objects: Early visual processing of spatial form defined by luminance, color, texture, motion, and binocular disparity. Sunderland, MA: Sinauer.