COGSCI 200 Exam 1 UPDATED ACTUAL Questions
and CORRECT Answers
Marr's Three Levels of Explanation: Functional - defines the problem
What are the levels? Algorithmic - describes the procedures or rules
Physical - identifies the neural/chemical substrates
"Inner Picture" Theory of (Visual) No
Perception: Is it right?
"Inner Picture" Theory of (Visual) Critical logical problem posed by the sources:
Perception: What objection(s) have "who is 'perceiving' the inner picture?"
been leveled against it? (Hint: think
homunculus)
"Inner Picture" Theory of (Visual) The "Inner Picture" theory of visual perception
Perception: What does it have to do posits that the brain operates like a camera or
with Forward Optics? projector, creating a 2D, picture-like replica of the
external world inside the mind
Inverse Optics Problem: What is the How do we get from the input (2D pattern on the
Inverse Optics Problem? retina) to the output (a 3D scene).
Inverse Optics Problem: What is the Making an inference from a 3D scent to what 2D
Forward Optics Problem? pattern will end up on the retina
,Inverse Optics Problem: Why is the Inverse Optics (Hard, One-to-Many Mapping) The
Inverse Optics Problem so hard brain, however, must solve the reverse problem:
(especially compared to the Forward taking the ambiguous 2D retinal image and
Optics Problem)? Hint: Know what reconstructing the 3D reality. This is hard because
the words/phrases underdetermined, there is a one-to-many mapping between the input
underspecified, and one-to-many and the output. Any single 2D pattern of light on
mapping and one-to-one mapping the retina could theoretically be caused by an
mean infinite number of different 3D scenes
Vision as Inference: What is a method of reasoning about an uncertain world
probabilistic inference? that allows a system (such as the human brain) to
update its beliefs on the basis of new evidence
Vision as Inference: Understand why the brain applies hidden assumptions—also called
lots of probabilistic inference natural constraints or priors—to narrow down the
involves hidden assumptions possibilities.
Vision as Inference: Is vision a kind of Yes. Based on the sources, vision is fundamentally
probabilistic inference? a form of probabilistic inference.
Because the visual information provided by the
eyes is ambiguous, the brain cannot simply
"deduce" the state of the world; instead, it must
guess the most probable state of the world based
on the available evidence
, Vision as Inference: What are the 1. The Single Overhead Light Source Assumption
hidden assumptions that the brain The visual system assumes there is only one light
uses in visual perception (e.g., the source and that it comes from above
Occlusion Assumption)? 2. The Shadow (Lightness Constancy) Assumption
The brain assumes that "things that are in a shadow
are really lighter in color than what my sensors
currently detect"
3. The Straight Line Assumption
The visual system assumes that things tend to move
in a straight line
4. The Rigidity Assumption
The brain assumes that all points on a moving
object move in synchrony
5. The Occlusion Assumption
The brain assumes that a moving object will
progressively cover and uncover portions of a
persisting background
6. The "Generic Viewpoint" Assumption
(Coincidence Avoidance)
The visual system "abhors a coincidence" and
assumes that the observer is not viewing a scene
from a highly unlikely, accidental angle
and CORRECT Answers
Marr's Three Levels of Explanation: Functional - defines the problem
What are the levels? Algorithmic - describes the procedures or rules
Physical - identifies the neural/chemical substrates
"Inner Picture" Theory of (Visual) No
Perception: Is it right?
"Inner Picture" Theory of (Visual) Critical logical problem posed by the sources:
Perception: What objection(s) have "who is 'perceiving' the inner picture?"
been leveled against it? (Hint: think
homunculus)
"Inner Picture" Theory of (Visual) The "Inner Picture" theory of visual perception
Perception: What does it have to do posits that the brain operates like a camera or
with Forward Optics? projector, creating a 2D, picture-like replica of the
external world inside the mind
Inverse Optics Problem: What is the How do we get from the input (2D pattern on the
Inverse Optics Problem? retina) to the output (a 3D scene).
Inverse Optics Problem: What is the Making an inference from a 3D scent to what 2D
Forward Optics Problem? pattern will end up on the retina
,Inverse Optics Problem: Why is the Inverse Optics (Hard, One-to-Many Mapping) The
Inverse Optics Problem so hard brain, however, must solve the reverse problem:
(especially compared to the Forward taking the ambiguous 2D retinal image and
Optics Problem)? Hint: Know what reconstructing the 3D reality. This is hard because
the words/phrases underdetermined, there is a one-to-many mapping between the input
underspecified, and one-to-many and the output. Any single 2D pattern of light on
mapping and one-to-one mapping the retina could theoretically be caused by an
mean infinite number of different 3D scenes
Vision as Inference: What is a method of reasoning about an uncertain world
probabilistic inference? that allows a system (such as the human brain) to
update its beliefs on the basis of new evidence
Vision as Inference: Understand why the brain applies hidden assumptions—also called
lots of probabilistic inference natural constraints or priors—to narrow down the
involves hidden assumptions possibilities.
Vision as Inference: Is vision a kind of Yes. Based on the sources, vision is fundamentally
probabilistic inference? a form of probabilistic inference.
Because the visual information provided by the
eyes is ambiguous, the brain cannot simply
"deduce" the state of the world; instead, it must
guess the most probable state of the world based
on the available evidence
, Vision as Inference: What are the 1. The Single Overhead Light Source Assumption
hidden assumptions that the brain The visual system assumes there is only one light
uses in visual perception (e.g., the source and that it comes from above
Occlusion Assumption)? 2. The Shadow (Lightness Constancy) Assumption
The brain assumes that "things that are in a shadow
are really lighter in color than what my sensors
currently detect"
3. The Straight Line Assumption
The visual system assumes that things tend to move
in a straight line
4. The Rigidity Assumption
The brain assumes that all points on a moving
object move in synchrony
5. The Occlusion Assumption
The brain assumes that a moving object will
progressively cover and uncover portions of a
persisting background
6. The "Generic Viewpoint" Assumption
(Coincidence Avoidance)
The visual system "abhors a coincidence" and
assumes that the observer is not viewing a scene
from a highly unlikely, accidental angle
