Reminder: extra credit experiments .

Reminder: extra credit experiments

www.tatalab.ca

Reminder: extra credit experiments

www.tatalab.ca

Upcoming Reading

• Vokey and Read – Subliminal Messages – Tuesday next week

• Visual Search: finding a single item in a cluttered visual scene

Visual Search

• Visual Search: finding a single item in a cluttered visual scene

Visual Search

• Is there a green square?

Visual Search

• Is there a green square?

Visual Search

• Parallel search: like many independent spotlights

Visual Search

• Serial search: each item is selected until target is found

Visual Search

• Serial search: each item is selected until target is found

Visual Search

• Serial search: each item is selected until target is found

Visual Search

• Serial search: each item is selected until target is found

Visual Search

• How could you test which kind of search was happening?

Visual Search

• Search Slope: How long per item?

Visual Search

• Search Slope: How long per item?

Visual Search

• Search Slope: How long per item?

Visual Search

• Parallel search - search time is independent of distracter number

Search Slope

0

200

400

600

800

1000

1200

1400

1600

1800

2000

0 10 20 30 40 50 60

Distractors

Response Time (ms)

Visual Search

• Conjunction search: NOT FLAT!

Visual Search

• Serial Search - linear increase in search time with number of distractors

Search Slope

0

200

400

600

800

1000

1200

1400

1600

1800

2000

0 10 20 30 40 50 60

Distractors

Response Time (ms)

Visual Search

• Search slope for shape singletons is flat. What does this tell us about shape and attention?

Visual Search

• Search slope for color singletons is flat. What does this tell us about color and attention?

Visual Search

• Search Slopes can be flat for targets defined by:– color– orientation– curvature– motion– depth

• What does this imply about these features ?• What does it tell us about conjunctions of

features ?

Features and Objects in Visual Processing

The Visual World is an Arrangement of Features

• Color• Motion• Form• Depth• Orientation

Pre-attentive vs. Attentive Processing

• Pre-attentive processing– Does the visual system register some basic

features automatically (without attention)– if so, what features?– How would you know?

Pre-attentive vs. Attentive Processing

• Indicators of Pre-attentive processing

– 1. processing precedes orienting - if you shift your attention to something or someplace because of some processing you did on the information there, you must have done that processing without attending

Pre-attentive vs. Attentive Processing

• Indicators of Pre-attentive processing

– 2. processing done in parallel - if you can process features of several objects simultaneously, you must have done that processing without attention

Parts vs. Wholes

– We see wholes, but the visual system initially sees parts (i.e. features) of objects

Parts vs. Wholes

• For example:

We see two rectangles, but the visual pathways initially detects small lines with some orientation

Parts vs. Wholes

• Simple features form boundaries

We see two rectangles, but the visual pathways initially detects small circles with some color

Parts vs. Wholes

• Conjunctions don’t form boundaries

We see only one rectangle (at least initially) because the boundaries of the inner one are made of conjunctions – these require attention to be perceived

“Early parsing of the visual field is mediated by separate properties, not by particular combinations of properties”

What does Treisman conclude from this observation?

• “Analysis of properties and parts precedes their synthesis”

• What is the “strong prediction” Treisman makes?

Illusory Conjunctions

• “errors of synthesis”

Illusory Conjunctions

• Identify the letter on the left of the screen and the digit on the right

Illusory Conjunctions

Q 4

Illusory Conjunctions

What colored shapes did you see?

Illusory Conjunctions

• Illusory conjunction - when perceived combination of attributes was not present

Illusory Conjunctions

• Illusory conjunction - when perceived combination of attributes was not present

• Supports notion that primitive features are processed independently and then bound together to form objects

• This binding process is thought to require attention focused on the location of the object to be bound

Treisman’s Feature Integration Theory

•Early visual system parses scene into features represented in “feature maps”

•“Attention Spotlight” can be moved across an overlay of these feature maps

•Focused attention is required to “bind” features together into objects

Next Time: Attention and Consciousness

• Sensory information must be attended for it to be entered into awareness