The world as we see it….
….. so far.
The Neural Picture
Rods and cones
Ganglion cells (M cells, P cells)
LGN (
magnocellular
tract,
parvocellular
tract)
PVC (striate cortex)
Simple, complex,
hypercomplex
)
The world as we see it….
….. so far.
The Visual Picture
•
Simple features
•
Low level vision
Affordances….
What information is important to glean from the
environment?
How do I learn about the objects and events in
the environment that I inhabit?
ASSIGNMENT FOR NEXT CLASS MEETING
Machine Vision
Limits of machine vision
Limits of simple features
Limits of low
-
level vision
Mid level vision
What intermediate processes intervene between
the collection of basic features and the
understanding of objects?
How do we “parse” a scene?
AI definition
–
interpreter, compiler of data in a
context of ambiguity
Top down, bottom up types
Why is the visual scene not
like a bar code?
Stimulus variation
Differences in illumination
Multiple edges, lines
Incomplete boundaries
Multiple interpretations
Observer variation
Viewpoint
–
accidental or otherwise
How do we begin to extract meaning
from the messy visual array?
How do we use and choose from the features at
hand?
On what basis do we “decide” what
interpretation to “go with”
A final comparison of machine and
man (and, of course, woman too)
Effort
Acquisition of skill
Flexibility of the rules
Pragnanz
What is available in the
stimulus?
Edges
Textures
Figure
-
ground
Parts and wholes
Edge detection
Sum up the simple features?
___ + ___ + ___ = _________
Problems
Variation in illumination
Incomplete forms
Illusory contours
Gestalt rules of
organization
A bit of history…..
Rules of organization
Good continuation
Edge detection
Gestalt rule of good continuation
Edge detection
Gestalt rule of good continuation
Gestalt rule of occlusion
Edge detection
Gestalt rule of good continuation
Gestalt rule of occlusion
Closure
Perceptual Committees
Good continuation VS Occlusion
…..
a
nd the winner is?
What is available in the
stimulus?
Edges
Textures
Figure
-
ground
Parts and wholes
Texture Segregation
(texture
segmentation)
Similarity
Texture Segregation
(texture
segmentation)
Similarity
Proximity
Texture Segregation
(texture
segmentation)
Similarity
Proximity
Parallelism and Symmetry
Texture Segregation
(texture
segmentation)
Similarity
Proximity
Parallelism and Symmetry
Common region
Texture Segregation
(texture
segmentation)
Similarity
Proximity
Parallelism and Symmetry
Common region
Connectedness
What is available in the
stimulus?
Edges
Textures
Figure
-
ground
Parts and wholes
Figure Ground
What is the stimulus?
What is the background?
Surroundedness
Size
Symmetry
Parallelism
Extremal
eges
–
shading
Relative motion
Extremal
edges
Enter the password to open this PDF file:
File name:
-
File size:
-
Title:
-
Author:
-
Subject:
-
Keywords:
-
Creation Date:
-
Modification Date:
-
Creator:
-
PDF Producer:
-
PDF Version:
-
Page Count:
-
Preparing document for printing…
0%
Σχόλια 0
Συνδεθείτε για να κοινοποιήσετε σχόλιο