classifyimages-2012
TRANSCRIPT
-
7/21/2019 Classifyimages-2012
1/24
Classifying imagesD.A. Forsyth
-
7/21/2019 Classifyimages-2012
2/24
Image classification - features
Issue:
category will not produce a single, simple pattern
but it might have components that are distinctive, but move around
Idea:
look for distinctive local patches (visual words)
build a histogram
-
7/21/2019 Classifyimages-2012
3/24
Important trick: K-Means
Choose a fixed number of clusters
Choose cluster centers and point-cluster allocations to
minimize error
cant do this by search
there are too many possible allocations.
Algorithm
fix cluster centers; allocate points to closest cluster
fix allocation; compute best cluster centers
x could be any set of features for which we can compute a distance
(careful about scaling)
xj" i
2
j#elements of i'th cluster
$%
&'
(
)*i#clusters
$
-
7/21/2019 Classifyimages-2012
4/24
K-means
-
7/21/2019 Classifyimages-2012
5/24
Building visual words - I
Learn a dictionary
cluster patch representations with k-means
k will be big (1000s-100,000s)
-
7/21/2019 Classifyimages-2012
6/24
Building visual words - II
Encode an image
find all interest points
for each patch around each interest point
map patch to closest cluster center
build histogram of interest points
-
7/21/2019 Classifyimages-2012
7/24
Visual words
-
7/21/2019 Classifyimages-2012
8/24
Visual words
-
7/21/2019 Classifyimages-2012
9/24
Visual words
-
7/21/2019 Classifyimages-2012
10/24
Features from visual words
Histogram
good summary of what is in image; quick and efficient
insensitive to spatial reorganization
Spatial pyramid
build histograms of local blocks at various scales
less insensitive to spatial reorganization
-
7/21/2019 Classifyimages-2012
11/24
Spatial pyramids
-
7/21/2019 Classifyimages-2012
12/24
Some standard tasks
Caltech 101, 256
images of isolated objects in 101, 256 categories
Imagenet
same, 1000s of categories
Pascal image classification
not isolated, fewer categories
Scenes
eg indoor, etc.
Materials
-
7/21/2019 Classifyimages-2012
13/24
Evaluation methods
Total error rate percentage of classification attempts that get the wrong answer
Accuracy
percentage of classification attempts that get the right answer
Class confusion matrix
table showing how classes are mixed up
Look at errors
-
7/21/2019 Classifyimages-2012
14/24
Accuracy on Caltech
-
7/21/2019 Classifyimages-2012
15/24
Material classification
-
7/21/2019 Classifyimages-2012
16/24
Spatial pyramids + scenes
-
7/21/2019 Classifyimages-2012
17/24
Evaluation
Precision
percentage of items in retrieved set that are relevant
Recall
percentage of relevant items that are retrieved
Precision vs recall
use classifier to label a collection of images
now plot precision against recall for different classifier thresholds
AP average precision
average of precision as a function of recall
-
7/21/2019 Classifyimages-2012
18/24
-
7/21/2019 Classifyimages-2012
19/24
Precision vs recall
-
7/21/2019 Classifyimages-2012
20/24
-
7/21/2019 Classifyimages-2012
21/24
Can be hard
-
7/21/2019 Classifyimages-2012
22/24
-
7/21/2019 Classifyimages-2012
23/24
-
7/21/2019 Classifyimages-2012
24/24