Miriah Meyer University of Utah

cs6630 | September 2 2014

PERCEPTION Miriah Meyer University of Utah

1 cs6630 | September 2 2014

PERCEPTION Miriah Meyer University of Utah

slide acknowledgements: Hanspeter Pﬁster, Harvard University Bang Wong, Broad Institute 2 administrivia . . .

3 - design critiques start this week - please include the visualization in the post ! - ﬁrst assignment goes out Thursday

4 last time . . .

5 Tufte’s integrity principles

Clear, detailed, and thorough labeling should be used to defeat graphical distortion and ambiguity.

The representation of numbers, as physically measured on the surface of the graphic itself, should be directly proportional to the numerical quantities represented.

Show data variation, not design variation.

6 Tufte’s design principles

- maximize the data-ink ratio

- avoid chart junk (sometimes)

- use multifunctioning elements

- layer information

- maximize the data density - shrink the graphics - maximize the amount of data shown (sometimes)

7 8 http://www.mikemake.com/72772/Charting-the-Beatles today . . .

9 - the eye

- edge detection

- relativity of perception

- things that pop

- gestalt principles

10 - the eye

- edge detection

- relativity of perception

- things that pop

- gestalt principles

11 12 120 million rods 5-6 million cones

Wandell, “Foundations of Vision” (left) 13 David R. Williams, Univ. of Rochester (right) Cone Response

14 HyperPhysics, Georgia State University 15 Ware 2010 16 C. Ware, “Visual Thinking for Design” E. B. Goldstein “Sensation and Perception” 17 (Adapted from Lindsay & Norman, 1977) Saccadas

- rapid involuntary eye movements - moving: 20-100 ms - ﬁxations: 200-600 ms

18 http://vision.arc.nasa.gov/personnel/jbm/home/projects/osa98/osa98.html Saccadas

- rapid involuntary eye movements - moving: 20-100 ms - ﬁxations: 200-600 ms

18 http://vision.arc.nasa.gov/personnel/jbm/home/projects/osa98/osa98.html - the eye

- edge detection

- relativity of perception

- things that pop

- gestalt principles

19 center surround

20 receptive ﬁeld

100M rods and cones

1M ganglion cells

Bear, Connors, Paradiso, “Neuroscience” on-center off-center

retinal ganglion cells

source: wikipedia low activity activity increased center and surrounds cancel or decreased at edges

Hermann grid eﬀect

Annie Prud’homme-Généreux

Takeaway

Our visual system sees differences, not absolute values, and is attracted to edges.

! Maximize the contrast with the background if the outlines of shapes are important. consequences of edge extraction

28 Cornsweet Illusion Cornsweet Illusion Cornsweet Illusion Cornsweet Illusion

D. Purves and R. B. Lotto Cornsweet Illusion

D. Purves and R. B. Lotto

Contrast Sensitivity

C. Ware, “Visual Thinking for Design” Contrast Sensitivity

C. Ware, “Visual Thinking for Design” Takeaway

The brain constructs surface color based largely on edge contrast information.

! We have higher contrast sensitivity in the luminance than in the chrominance channel. - the eye

- edge detection

- relativity of perception

- things that pop

- gestalt principles

38 WEBER’S LAW we judge based on relative, not absolute, differences

39 RELATIVE DIFFERENCES

40 AXIS OF ALIGNMENT

41 AXIS OF ALIGNMENT

41 SIMULTANEOUS CONTRAST

42 http://persci.mit.edu/_media/gallery/checkershadow_double_full.jpg SIMULTANEOUS CONTRAST

43 SIMULTANEOUS CONTRAST

INTERACTION OF COLOR

46 INTERACTION OF COLOR

47 INTERACTION OF COLOR

48 Wong 2010 - the eye

- edge detection

- relativity of perception

- things that pop

- gestalt principles

49 POPOUT

50 POPOUT

50 Based on slide from Mazur PRE-ATTENTIVE PROCESSING

- requires attention, despite name

- very fast: <200 ms

- what matters most is contrast between features

52 BASIC POPOUT CHANNELS

53 Ware 2008 which side has the outlier? Healy 2007 Healy 2007 57 Healey 2007 CONJUNCTION or, why to use a single channel at a time

57 Healey 2007 CONJUNCTION objects to be searched

corresponding feature space

58 59 Takeaway

We can easily see objects that are different in color and shape, or that are in motion.

! Use color and shape sparingly to make the important information pop out. - the eye

- edge detection

- relativity of perception

- things that pop

- gestalt principles

61 Gestalt principles

- German: “Gestalt” = form

- patterns transcend the visual stimuli that produced them similarity

Bang Wong, “Gestalt Principles, 1”, Nature Methods similarity

Andy Rutledge, “Gestalt Principles of Perception” proximity

Bang Wong, “Gestalt Principles, 1”, Nature Methods connectedness

Ware, “Information Visualization” grouping

Similarity

Connection

Enclosure

Bang Wong, “Gestalt Principles, 1”, Nature Methods grouping

Bang Wong, “Gestalt Principles, 1”, Nature Methods grouping proximity similarity

enclosure

Jorge Camoes continuity

Bang Wong, “Gestalt Principles, II”, Nature Methods continuity

Bang Wong, “Gestalt Principles, II”, Nature Methods closure closure PATHLINE A TOOL FOR COMPARATIVE FUNCTIONAL GENOMICS PATHWAY METRIC OVERVIEW SPECIES CURVEMAP OVERLAYS g1 g2 PATHLINE-0.50 1.00 A TOOL FOR COMPARATIVE FUNCTIONAL GENOMICS 7.7 PATHWAY METRIC OVERVIEW SPECIES CURVEMAP OVERLAYS P1 m1 s1 g1 g2 g1 g2 -7.1 PATHLINE-0.50 1.00 A TOOL FOR COMPARATIVE FUNCTIONAL GENOMICS m2 7.7 8.3 PATHWAY METRIC OVERVIEWa SPECIES CURVEMAP OVERLAYS P1 m1 s1 s2 g1 g2 g1 g2 -7.1 -5.0 PATHLINE-0.50m2 1.00 A TOOL FOR COMPARATIVE FUNCTIONAL GENOMICS 7.7 8.3 5.4 PATHWAY METRIC OVERVIEWa SPECIES CURVEMAP OVERLAYS P1 m1

B1 s1 s2 s3 b g1 g2 g1 g2 -7.1 -5.0 -3.8 PATHLINE-0.50 1.00 A TOOL FOR COMPARATIVE FUNCTIONAL GENOMICS m2 7.7 8.3 5.4 3.0 PATHWAY METRIC OVERVIEWa SPECIES CURVEMAP OVERLAYS P1 m1 B1 s1 s2 s3 s4 b g1m7 g2 g1 g2 -7.1 -5.0 -3.8 -5.2 c PATHLINE-0.50 1.00 A TOOL FOR COMPARATIVE FUNCTIONAL GENOMICS m2 7.7 8.3 5.4 3.0 4.6 PATHWAY METRIC OVERVIEWa SPECIES CURVEMAP OVERLAYS P1 m1 d s4 B1 s1 s2 s3 s5 b g1m7 g2 g1 g2closure -7.1 -5.0 -3.8 -5.2 -4.4 c m10 PATHLINE-0.50 1.00 A TOOL FOR COMPARATIVE FUNCTIONAL GENOMICS m2 7.7 8.3 5.4 3.0 4.6 4.2 PATHWAY METRIC OVERVIEWa SPECIES CURVEMAP OVERLAYS

P1 m1 d s5 s6 B1 s1 s2 s3 s4 b g1m7 g2 g1 g2 -7.1 -5.0 -3.8 -5.2 -4.4 0.0 c m10 -0.50 1.00 m2 m10 7.7 8.3 5.4 3.0 4.6 4.2 2.3 a P1 m1 d s6 s7 B1 s1 s2 s3 s4 s5 b g1m7 g2 -7.1 -5.0 -3.8 -5.2 -4.4 0.0 -3.4 c m10 m2 m10 8.3 5.4 3.0 4.6 4.2 2.3 3.9 a d s2 s4 s5 s6 s7 s8 B1 s3 b m7 -5.0 -3.8 -5.2 -4.4 0.0 -3.4 -1.5 c m10 m10 5.4 3.0 4.6 4.2 2.3 3.9 4.7 s8 B1 d s3 s4 s5 s6 s7 s9 b m7 m18

P2 -3.8 -5.2 -4.4 0.0 -3.4 -1.5 -2.1 c m10 m10 3.0 4.6 4.2 2.3 3.9 4.7 3.7 d s4 s5 s6 s7 s8 s9 s10 m7 m18 g30

P2 -5.2 -4.4 0.0 -3.4 -1.5 -2.1 -4.6 c m10 m10 4.6 4.2 2.3 3.9 4.7 3.7 3.9 d s5 s6 s7 s8 s9 s10 s11 m18 g30

P2 -4.4 0.0 -3.4 -1.5 -2.1 -4.6 -3.6 m10 m10 4.2 2.3 3.9 4.7 3.7 3.9 0.5 s6 s7 s8 s9 s10 s11 s12 m18 g30

P2 0.0 -3.4 -1.5 -2.1 -4.6 -3.6 -2.6 m10 2.3 3.9 4.7 3.7 3.9 0.5 0.0 s7 s8 s9 s10 s11 s12 s13 m18 g30

P2 -3.4 -1.5 -2.1 -4.6 -3.6 -2.6 -1.1 -0.50 1.00 3.9 4.7 3.7 3.9 0.5 0.0 1.3 s8 s9 s10 s11 s12 s13 s14 g30 KEY Genes m18 -1.1 -1.0 P2 -1.5 -2.1 -4.6 -3.6 -2.6 forward reverse bidirectional m28 -0.50 1.00 4.7 3.7 3.9 0.5 0.0 1.3 Metabolites Metrics s14 8.3 4.7 KEY Genes m2 s9 s10 s11 s12 s13 m18 g30 P3 PearsonSubgroup1

P2 -2.1 -4.6 -3.6 -2.6 -1.1 -1.0 forward reverse bidirectional m28 PearsonSubgroup2 -0.50 1.00 3.7 3.9 0.5 0.0 1.3 Metabolites m33Metrics 8.3 4.7 s10 PearsonALLs12 s13 s14 -5.2 -7.1 KEY Genes m2 s11 g30 P3 PearsonSubgroup1 m13 -4.6 -3.6 -2.6 -1.1 -1.0

forward reverseP4 bidirectional m28 PearsonSubgroup2 -0.50 1.00 3.9 0.5 0.0 1.3 m33 8.3 4.7 Metabolites Metrics PearsonALLs11 s13 s14 -5.2 -7.1 KEY Genes m2 s12 P3 PearsonSubgroup1 m13 m30 -3.6 -2.6 -1.1 -1.0

forward reverseP4 bidirectional m28 PearsonSubgroup2 -0.50 1.00 0.5 0.0 1.3 m33 8.3 4.7 Metabolites Metrics PearsonALLs12 s14 -5.2 -7.1 KEY Genes m2 s13 P3 PearsonSubgroup1 m13 m30 -2.6 -1.1 -1.0

forward reverseP4 bidirectional m28 PearsonSubgroup2 -0.50 1.00 0.0 1.3 Metabolites m33Metrics 8.3 4.7 PearsonALL s13 s14 -5.2 -7.1 KEY Genes m2 P3 PearsonSubgroup1 m13 m30 -1.1 -1.0

forward reverseP4 bidirectional m28 PearsonSubgroup2 -0.50 1.00 1.3 m33 8.3 4.7 Metabolites Metrics PearsonALL s14 -5.2 -7.1 KEY Genes m2 P3 PearsonSubgroup1 -1.0 forward reverse bidirectionalm13 m30 m28 P4 PearsonSubgroup2 Metabolites m33Metrics PearsonALL 8.3 4.7 m2 -5.2 -7.1 P3 PearsonSubgroup1 m13 m30 P4 PearsonSubgroup2 m33 PearsonALL -5.2 -7.1 m13 m30 P4

m30 ﬁgure / ground alignment

Bang Wong, “Gestalt Principles, II”, Nature Methods common fate

A. Rutledge

Gestalt principles - similarity: things that look like each other (size, color, shape) are related

- proximity: things that are visually close to each other are related

- connection: things that are visually connected are related

- continuity: we complete hidden objects into simple, familiar shapes

- closure: we see incomplete shapes as complete

- ﬁgure / ground: elements are perceived as either ﬁgures or background

- common fate: elements with the same moving direction are perceived as a unit

L4. Data abstraction REQUIRED READING

82 83