Intro to Vision Science Terms

Question 1

Visual Illusions

Answer 1

misperceptions due to processing in visual system, result in a percept that differs from reality (NOT OPTICAL ILLUSIONS)

eg. apparent convergence of vertical lines - rising objects appear to fall outward at the top - connects to buildings with columns (parthenon used distorted building made them inclined inwards to compensate for visual illusion and make them appear parallel and straight

Question 2

Sensation

Answer 2

simple sensory experience elicited by a stimulus; conversion of energy or chemicals to a neural code

eg. senses are vision, audition, gustation, touch, warmth, cold, pain

Question 3

Information Processing

Answer 3

describes the behavior of a particular kind of device (info processor),

looking at mind as an info processor which is a complex system that receives, stores, retrieves, transforms and transmits information, done by processing info (data) according to rules/procedures (program).

3 accounts: implementational, formal, interpretational

Question 4

Perception

Answer 4

conscious experience of objects; organization and interpretation of neural signals

eg. recognition or identification of an object based on sensations, using vision to identify the lightwaves/shapes that make up a phone as a phone

Question 5

Implementational Aspects of Perception

Answer 5

how perceptual procedures are physically realized in the “wetware” of the brain, described in terms of physical structures involved, neurophysiology aspect

Question 6

Formal Aspects of Perception

Answer 6

information processes are represented in symbols and abstract structures, described in terms of underlying functions and knowledge structures, rules are formal only depend on token type

Question 7

Interpretational Aspects of Perception

Answer 7

a description of what a system is doing and why, in terms of the behaviors observed, what is the goal

Question 8

Tri-Level Hypothesis

Answer 8

basic assumption is perception is information processing, 3 levels

1) Computational theory (behavioral level) - goal of computation, why appropriate, logic strategy by which it can be carried out, logic/mathematics.

2) representation/algorithm (functional level), how can theory be actualized, what is representation for input/output, what is algorithm for transformation psychophysics/computer simulation

3) Hardware implementation (physical level), how can representation and algorithm be realized physically, neuroscience/electrical engineering

Question 9

Primitive Operations

Answer 9

makes up an info processing system, makes up the functional architecture, biologically fixed part of cognition,

eg. analogy is the dictionary from which all sentences in some language can be created

Question 10

Functional Analysis

Answer 10

take a very complicated functional description and decompose it into organized description of simpler subfunctions, typically just leads to descriptions not to explanations

eg. Ryles Regress: colour perception -> processing of info in colour pathways -> neural circuits -> neuroanatomy -> neurobiochemistry -> chemistry -> physics -> ?

Question 11

Subsumption Criterion

Answer 11

if a sub function can be explained using natural, causal laws (physiological laws) then decomposition can stop

eg. colour vision was described functionally in trichromatic theory but was not subsumed until three different photopigment molecules isolated in 1960s, there is no need to explain how / why molecules are photo sensitive

Question 12

Cognitive Penetrability Criterion

Answer 12

if a perceptual phenomenon cannot be altered by a change in beliefs, then the function is cognitively impenetrable, if it is not changed by top-down processing then it implies it is bottom-up and thus must be handled by the functional architecture

eg. visual illusion like the simultaneous colour contrast, even if you know what colours actually are (same shade of red) it does not change the experience of the illusion (one red seeming more orangeish and the other red seeming darker)

Question 13

Weakly Equivalent Models

Answer 13

from engineering perspective, just develop system that accomplishes some task not necessarily in the same way as humans (pure AI), system can pass a test of intelligence for wrong reasons from a psychological perspective. Same computation, different representation/algorithm, different implementation

eg. AI playing chess

Question 14

Strongly Equivalent Models

Answer 14

from cognitive science perspective, want to develop models of human processes that exhibit the same behavior for the right reasons (same underlying fundamental processes being used). Same computation, same representation/algorithm, different implementation