Please enable JavaScript.
Coggle requires JavaScript to display documents.
Recognizing Objects (Word recognition (Feature nets (Middle: Perceiving…
Recognizing Objects
Word recognition
Familiarity
More common words are easier to recall
Recency
primed words were recognized
Word superiority
Words are easier to perceive than individual letters
Feature nets
Middle: Perceiving the letters themselves
Bottom: Detectors for simple features of each letter
Top: The word itself
Bigram detector
Added layer to perceive letter combinations
Object Recognition
Features called “geons”
Recognition by Component
Features --> geon detectors --> geon assemblies
Multiple Views
Viewpoint independent: can be recognized from almost any angle
Viewpoint dependent: Recognition is slower from some angles
Facial recognition
Faces do not seem to be constructed of geons
Faces are dependent on viewpoint and orientation
Propoganosia
Inability to recognize faces
Recognizing faces depends on holistic recognition
Basic recognition
begins with features of an object
Common characteristics of objects
Single feature vs. combination
More difficult for a combination of features
Types of processing
Bottom-up: Data driven/simple features first
Top-down: Concept driven/knowledge first