visual cognition ii object perception. theories of object recognition template matching models...

38
Visual Cognition II Object Perception

Post on 19-Dec-2015

231 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Visual Cognition II

Object Perception

Page 2: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Theories of Object Recognition

• Template matching models• Feature matching Models• Recognition-by-components• Configural models

Page 3: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Template matching

Detect patterns by matching visual input with a set of templates stored in memory – see if any template matches.

TEST INSTANCE

“J” TEMPLATE “T” TEMPLATE

match

Page 4: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Problem:

what if the object differs slightly from the template? E.g., it is rotated or scaled differently?

Solution:

use a set of transformations to best align the object with a template (using translation, rotation, scaling)

TEST INSTANCE

“J” TEMPLATE “T” TEMPLATE

rotation

match

Page 5: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Template-matching works well in constrained environments

Page 6: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Figure 2-15 (p. 58)Examples of the letter M.

Problem: template matching is not powerful enough for general object recognition

Page 7: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Feature Theories

• Detect objects by the presence of features• Each object is broken down into features• E.g.

A = + +

Page 8: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Problem

• Many objects consist of the same collection of features

• Need to also know how the features relate to each other structural theories

• One theory is recognition by components

Different objects, similar sets of features

Page 9: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Recognition by Components (RBC)

• Biederman (1987): Complex objects are made up of arrangements of basic, component parts: geons.

• “Alphabet” of 24 geons

• Recognition involves recognizing object elements (geons) and their configuration

Page 10: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Why these geons?

• Choice of shape vocabulary seems a bit arbitrary• However, choice of geons was based on non-accidental

properties. The same geon can be recognized across a variety of different perspectives:

except for a few “accidental” views:

Page 11: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Viewpoint Invariance

• Viewpoint invariance is possible except for a few accidental viewpoints, where geons cannot be uniquely identified

Page 12: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Prediction

• Recognition is easier when geons can be recovered

• Disrupting vertices disrupts geon processing more than just deleting parts of lines

ObjectDeleting

line segments

Deleting vertices

Page 13: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Evidence from priming experiments

Page 14: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Problem for RBC

• Theory does not say how color, texture and small details are processed. These are often important to tell apart specific exemplars or similar objects. E.g.:

Page 15: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Configural models of recognition

• Individual instances are not stored; what is stored is an “exemplar” or representative element of a category

• Recognition based on “distance” between perceived item and prototype

prototype

match

“Face space”

no match

Page 16: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Prediction: Caricatures might be better recognized than original face

from: Face Recognition by Humans: 20 Results all Computer Vision Researchers Should Know About. Sinha et al. (2005).

average female face “veridical” face caricature of B

Page 17: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

a) no, nothingb) same mouthc) same nosed) same eyes

Do these faces have anything in common?

Page 18: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

How about these ones?

By disrupting holistic (configural) processing, it becomes easier to process the individual parts

Page 19: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

• Configural effects often disappear when stimulus is inverted

Face Inversion

Page 20: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Top-down and Context Effects in Object Recognition

Page 21: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Slide from Rob Goldstone

Page 22: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Context can often help in identification of an object

Later identification of objects is more accurate when object is embedded in coherent context

Page 23: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Context can alter the interpretation of an object

Page 24: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Context Effects in Letter Perception

The word superiority effect: discriminating between letters is easier in the context of a word than as letters alone or in the context of a nonword string.

DEMO:http://psiexp.ss.uci.edu/research/teachingP140C/demos/demo_wordsuperiorityeffect.ppt (Reicher, 1969)

Page 25: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Interactive Activation Model

• Word superiority effect suggests that information at the word level might affect interpretation at the letter level

• Interactive activation model: neural network model for how different information processing levels interact

• Levels interact– bottom up: how letters combine to form words– top-down: how words affect detectability of letters

Page 26: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

The Interactive Activation Model

• Three levels: feature, letter, and word level

• Nodes represent features, letters and words; each has an activation level

• Connections between nodes are excitatory or inhibitory

• Activation flows from feature to letter to word level and back to letter level

(McClelland & Rumelhart, 1981)

Page 27: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

The Interactive Activation Model

• Bottom-up:– feature to word level

• Top-down: – word to letter level

• Model predicts word superiority effect because of top-down processing

(McClelland & Rumelhart, 1981)

Page 28: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Predictions of the IA model – stimulus is “WORK”

• At word level, evidence for “WORK” accumulates over time• Small initial increase for “WORD”

WORK

WORD

WEAR

Page 29: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Predictions of the IA model – stimulus is “WORK”

Why does the letter “K” get activated?

a) because of (partial) activation from feature level

b) because of activation from word level back to feature level

c) both a) and b)

K

R

D

Page 30: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Predictions of the IA model – stimulus is “WORK”

Why does the letter “R” get partially activated?

a) because of (partial) activation from feature level

b) because of activation from word level back to feature level

c) both a) and b)

K

R

D

Page 31: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

For a demo of the IA model, see:

http://www.itee.uq.edu.au/~cogs2010/cmc/chapters/LetterPerception/

Page 32: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

“Mind reading”

Page 33: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Predicting What Somebody is Seeing (“mind reading”)

Viewing a Bottle Viewing a Shoe

If the brain response is different for different kinds of stimuli, can we predict what somebody is thinking of solely based on the brain’s response?

bold response bold response

Page 34: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Pattern Classification Method

1. Acquire brain data for different stimuli (e.g. bottles and shoes)

2. Train a classifier (such as the neural network on right) to discriminate between bottle voxel patterns and shoe voxel patterns

3. Test classifier on novel images

(slide from Ken Norman)

bottle shoe

Input Layer (voxels)

Output layer(categories)

Page 35: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Faces

Cats

Scissors

Chairs

Houses

Bottles

Shoes

Scrambled Pictures

slides courtesy of Jim Haxby

Haxby et al. (2001)can predict with 96% accuracy stimuli from 8 categories

Page 36: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Reconstructing the Mental Image

• If we can predict what somebody is looking at, can we also reconstruct what somebody might be looking at from just the brain’s response?

Image Brain’s responseMathematical

Model Reconstructed image

Page 37: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Reconstructing simple patterns from fMRIMiyawaki et al. (2008)

from: Miyawaki et al. (2008). Neuron, 60(5), pp. 915-929. movie at: http://psiexp.ss.uci.edu/research/teachingP140C/demos/mmc2.mpg

Page 38: Visual Cognition II Object Perception. Theories of Object Recognition Template matching models Feature matching Models Recognition-by-components Configural

Brain Computer Interfaces

ATR Laboratories in Japan developed arobotic hand that can be controlledusing fMRI

Rainer Goebel’s team had two patients play mental ping-pong in fMRI machines