Form perception
1 Structuralism
Complex mental processes are created by combining fundamental components. Tichener, 19th Century. In other words, perception is an aggregate of simple elements. Eventually more than 40,000 "elementary sensations" were listed.
2 3 Gestalt ideas of form perception
Founder of Gestalt Psychology is Max Wertheimer
Wikipedia
4 Gestalt principles of organization (how figural properties determine perceived patterns):
grouping due to proximity
grouping due to similarity
grouping due to common motion
closure
figure-ground perception
A central tenet: The whole is different from the sum of its parts
5 Grouping by proximity
A B C
Image by MIT OpenCourseWare.
6 Grouping by shape
Image by MIT OpenCourseWare.
7 Three general theories of form perception:
1. Form perception is accomplished by neurons that respond selectively to line segmens of different orientation.
2. Form perception is accomplished by spatial mapping of the visual scene onto visual cortex.
3. Form perception is accomplished by virtue of Fourier analysis.
8 1. Orientation of line segments and spatial frequency
9 Form perception with little information about orientation of line segments
Image removed due to copyright restrictions.
10 Proof that low spatial-frequency information is used for face recognition
Image removed due to copyright restrictions.
11 Image removed due to copyright restrictions.
12 Image removed due to copyright restrictions.
13 2. Topographic mapping
14 extreme isomorphism
15 Cortical layout of neurons activated by arrows
7 90 6 45 5 4 3 2 1 all arrows in
visual field 0 0 contralateral hemifield
-45 -45
-90 -90
-45 monkey area V1 0
45 0.25 0.25 0.5 0.5 1 90 1 2 2 4 7 7 4
Image by MIT OpenCourseWare. 16 Cortical layout of neurons activated by arrows
7 90 6 45 5 45 4 3 2 arrows across visual field 1 midline 0 0
-45 -45
-90 -90
-45 monkey area V1 0
45 0.25 0.25 0.5 0.5 1 90 1 2 2 4 7 7 4
Image by MIT OpenCourseWare. 17 File: prosthetics14 Cortical layout of neurons activated by a circle
90 Visual field 90 5 5 135 4 45 135 4 45 3 3 2 2 1 1
180 0 180 0
225 315 225 315
270 270 Posterior view of monkey brain showing the striate cortex and activated regions
5 4 5 4 3 4 5 3 3 4 5 3 2 270 2 2 270 2 1 270 1 1 270 1
315 225 315 225
0 180 0 180
45 135 45 135
90 90 90 90
left hemisphere 10 mm right hemisphere left hemisphere 10 mm right hemisphere
Figure 5B, C. Schiller, Peter H., and Edward J. Tehovnik. "Visual Prosthesis." Perception 37, no. 10 (2008): 1529. Courtesy of Pion Ltd. Used with permission.
18 The Giotto Story:
When Pope Benedict, the 12th, in 13th Century set out to have the walls of the great cathedral of St. Peter redecorated, he sent messengers all over Italy to find out who were the best painters. Specimen were gladly given. When a messenger came to Ambrogio Bondone Giotto (1267-1337), he did not provide a sample painting. Instead, he took a sheet of paper and a pencil dipped in red color, and drew a perfect circle. "Here is your drawing," he said. The Pope, upon examining all the productions submitted, chose Giotto without hesitation.
To this day in Tuscany there is the saying:
The round O of Giotto
19 C D A B
G H E F
I K M L
Image by MIT OpenCourseWare.
20 3. Spatial frequency analysis
Fergus Campbell and John Robson
21 deLange functions
Spatial Frequency (cycles/degree)
0.1 1 10 100
0.001 1,000 INVISIBLE Sensitivity(1/thresholdcontrast)
0.01 100
VISIBLE 0.1 10 Threshold Contrast
1.0 1
0.1 1 10 100 Spatial Frequency (cycles/mm on retina)
Image by MIT OpenCourseWare. 22 Generating compound gratings from a set of simple ones
simple gratings compound graings compound graings at low contrast
23 Frequency-specific adaptation
24 Frequency-specific adaptation
Contrast Sensitivity(1/ThresholdContrast) 1,000 1,000
100 100
10 10
Contrast Sensitivity (1/Threshold Contrast) Contrast Sensitivity 1 1
Spatial Frequency
Frequency specific adaptation
Image by MIT OpenCourseWare.
25 Channel model
0.001 Sensitivity of a 1,000 Single Channel Contrast Sensitivity(1/thresholdcontrast)
0.01 100
Threshold Contrast 0.1 10
1 1 0.1 1 10 100 Spatial Frequency (cycles/degree)
CHANNEL MODEL Image by MIT OpenCourseWare. 26 Four essential properties of spatial frequency analysis:
spatial frequency
contrast
orientation
phase
When information about these four properties is available, one can reconstruct any visual pattern
27 Shape-selective responses in inferotemporal cortex
28 IT neuron response to various shapes
CELL 1
5o
A B C D
1 -.01* -.04* .44*
MO 03124 E F G H 20i/s
.09* .20* .31* .15* 5s
Image by MIT OpenCourseWare.
29 IT neuron response to various shapes
CELL1
5o
A B C D
1 .12* .32* .89*
SU 0162
E F G H .09* .03* .15* .36* 10i/s
5s
Image by MIT OpenCourseWare.
30 Image by MIT OpenCourseWare.
31 Intermediate level vision
32 Basic visual capacities color brightness pattern texture motion depth
Intermediate visual capacities constancy selection recognition transposition and invariance comparison location
33 Transposition and invariance
34 Hirschfeld
Image removed due to copyright restrictions.
35 Image removed due to copyright restrictions.
36 Image removed due to copyright restrictions.
Voltaire, 18th French Enlightment writer of more than 2,000 books and pamphlets
37 The Art World
Secret Knowledge by David Hockney
38 Masolino da Panicale, 1425 Robert Campin, 1430
Images are in the public domain. 39 Image by MIT OpenCourseWare.
Camera obscura
40 Jan van Eyck 1436
Painting is in public domain. 41 Bacchus, c1597 by Caravaggio
Camera obscura??? Painting is in public domain. David Hockney 42 original image camera obscura image
rotated camera obscura image
Painting is in public domain. 43 The marriage of Giovanni Arnolfini with Giovanna Cenami Jan van Eyck, 1434
C4amera obscura??? David Hockney Painting is in public domain.
44 The recognition of faces
45 The recognition of faces
Photos removed due to copyright restrictions.
46 The recognition of faces
Photos removed due to copyright restrictions.
47 The perception of subjective contours
48 49 50 51 A B Subjective contours at high contrast Subjective contours at isoluminance
Image by MIT OpenCourseWare.
52 Response to subjective contours in V2
UNIT 3GD5 +
A 27.7
B 4.6
Image by MIT OpenCourseWare.
53 Response to subjective contours in V2 A B
0.5s (20)
(A) (B)
Image by MIT OpenCourseWare.
54 The effects of V4 and MT lesions on intermediate vision
55 Match to sample task
56 The effect of V4 lesions on object matching
Figure removed due to copyright restrictions.
57 Intermediate vision tasks SIZE
A B C
OCCLUSION A B C
CONTOUR A B C
58 The effect of V4 lesions on object transformations
Figure removed due to copyright restrictions.
59 Greater and lesser brightness discrimination
Target Brighter Target D immer
60 Brightness discrimination, greater and lesser targets
Figure removed due to copyright restrictions.
61 Greater and lesser size discrimination
Target Larger Target S maller
62 Size discrimination, greater and lesser targets
Figure removed due to copyright restrictions.
63 Confusing and interesting percepts created by great artists
64 Courtesy of The M. C. Escher Company, B. V. Used with permission.
Maurits Cornelis Escher (b. 1898)
65 Courtesy of The M. C. Escher Company, B. V. Used with permission. 66 Courtesy of The M. C. Escher Company, B. V. Used with permission. 67 Summary, form:
1. Three theories of form precessing in the brain are (a) analysis by orientation of line segmens, (b) spatial mapping onto a topographically organized brain region and (c) Fourier analysis.
2. Areas V2, V4 and IT play important roles in intermediate vision.
3. Neurons responding to subjective contours have been found in V2.
4. Recognition of objects transformed in various ways is compromised by V4 and IT lesions. V4 lesions also produce major deficits in visual learning and in selecting "lesser" stimuli.
5. Some IT neurons are selective for objects including faces, but most respond to a variety of objects whose recognition is based on the differential activity of a great many neurons.
6. How we process and deal with ambiguities in perception remains a mystery
68 MIT OpenCourseWare http://ocw.mit.edu
9.04 Neural Basis of Vision and Audition Fall 2013
For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.