展开查看详情
1 . Object Recognition by Parts
• Object recognition started with line segments.
- Roberts recognized objects from line segments
and junctions.
- This led to systems that extracted linear features.
.
- CAD-model-based vision works well for industrial.
• An “appearance-based approach” was first developed
for face recognition and later generalized up to a point.
• The new interest operators have led to a new kind of
recognition by “parts” that can handle a variety of
objects that were previously difficult or impossible.
1
2 . Object Class Recognition
by Unsupervised Scale-Invariant Learning
R. Fergus, P. Perona, and A. Zisserman
Oxford University and Caltech
CVPR 2003
won the best student paper award
CVPR 2013
won the best 10-year award
2
3 . Goal:
• Enable Computers to
Recognize Different
Categories of Objects
in Images.
3
5 . Approach
• An object is a constellation of parts (from Burl, Weber
and Perona, 1998).
• The parts are detected by an interest operator (Kadir’s).
• The parts can be recognized by appearance.
• Objects may vary greatly in scale.
• The constellation of parts for a given object is learned
from training images
5
6 . Components
• Model
– Generative Probabilistic Model including
Location, Scale, and Appearance of Parts
• Learning
– Estimate Parameters Via EM Algorithm
• Recognition
– Evaluate Image Using Model and Threshold
6
7 . Model: Constellation Of
Parts
Fischler & Elschlager,
1973
Yuille, ‘91
Brunelli & Poggio, ‘93
Lades, v.d. Malsburg et al. ‘93
Cootes, Lanitis, Taylor et al. ‘95
Amit & Geman, ‘95, ‘99
Perona et al. ‘95, ‘96, ’98, ‘00
7
8 . Parts Selected by
Interest Operator
Kadir and Brady's Interest Operator.
Finds Maxima in Entropy Over Scale and Location
8
9 . Representation of
Appearance
c1
Projection onto c2
11x11 patch Normalize PCA basis
121 dimensions was too big, so they used PCA to reduce to 10-15.
c915
10 . Learning a Model
• An object class is represented by a generative
model with P parts and a set of parameters .
• Once the model has been learned, a decision
procedure must determine if a new image
contains an instance of the object class or not.
• Suppose the new image has N interesting
features with locations X, scales S and
appearances A.
10
11 . Probabilistic Model
• X is a description of the shape of the object (in terms of locations of parts)
• S is a description of the scale of the object
• A is a description of the appearance of the object
• θ is the (maximum likelihood value of) the parameters of the object
• h is a hypothesis: a set of parts in the image that might be the parts of the object
• H is the set of all possible hypotheses for that object in that image.
• For N features in the image and P parts in the object, its size is O(NP)
11
12 . Appearance
The appearance (A) of each part p Background model has mean cbg
has a Gaussian density with and covariance Vbg.
mean cp and covariance VP.
Gaussian Part Appearance PDF Guausian Appearance PDF
Object Background 12
13 . Shape as Location
Object shape is represented by a joint Gaussian density of the locations (X)
of features within a hypothesis transformed into a scale-invariant space.
Gaussian Shape PDF Uniform Shape PDF
Object Background 13
14 . Scale
The relative scale of each part is modeled by a Gaussian density with
mean tp and covariance Up.
Prob. of detection
Gaussian
Relative Scale PDF
0.8 0.75 0.9
Log(scale)
14
15 .Occlusion and Part Statistics
This was very complicated and turned out to not work
well and not be necessary, in both Fergus’s work and
other subsequent works.
15
16 . Learning
• Train Model Parameters
occlusion
Using EM: location
appearance scale
• Optimize Parameters
• Optimize Assignments
• Repeat Until Convergence
16
17 . Recognition
Make this likelihood ratio:
greater than a threshold.
17
18 . RESULTS
• Initially tested on the Caltech-4 data set
– motorbikes
– faces
– airplanes
– cars
• Now there is a much bigger data set: the
Caltech-101 http://
www.vision.caltech.edu/archive.html
18
19 .Equal error rate: 7.5% Motorbikes
19
20 .Background Images
It learns that these are NOT motorbikes.
20
21 .Equal error rate: 4.6% Frontal faces
21
22 .Equal error rate: 9.8%
Airplanes
22
23 .Scale-Invariant Cats
Equal error rate: 10.0%
23
24 . Scale-Invariant Cars
Equal error rate: 9.7%
24
25 . Accuracy
Initial Pre-Scaled Experiments
25
热门城市
