-
1
Component Analysis Methodsfor Computer Vision and
Pattern Recognition
Fernando De la TorreFernando De la Torre
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 1
Computer Vision and Pattern Recognition Easter School Computer
Vision and Pattern Recognition Easter School March 2011March
2011
Component Analysis for CV & PR • Computer Vision & Image
Processing
– Structure from motion.– Spectral graph methods for
segmentation.– Appearance and shape models.– Fundamental matrix
estimation and calibration.– Compression.– Classification.–
Dimensionality reduction and visualization.
• Signal Processing– Spectral estimation, system identification
(e.g. Kalman filter), sensor
array processing (e.g. cocktail problem, eco cancellation),
blind source
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 2
array processing (e.g. cocktail problem, eco cancellation),
blind source separation, …
• Computer Graphics– Compression (BRDF), synthesis,…
• Speech, bioinformatics, combinatorial problems.
• Computer Vision & Image Processing– Structure from
motion.– Spectral graph methods for segmentation.– Appearance and
shape models.
Structure from motion
Component Analysis for CV & PR
– Fundamental matrix estimation and calibration.– Compression.–
Classification.– Dimensionality reduction and visualization.
• Signal Processing– Spectral estimation, system identification
(e.g. Kalman filter), sensor
array processing (e.g. cocktail problem, eco cancellation),
blind source
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 3
array processing (e.g. cocktail problem, eco cancellation),
blind source separation, …
• Computer Graphics– Compression (BRDF), synthesis,…
• Speech, bioinformatics, combinatorial problems.
• Computer Vision & Image Processing– Structure from
motion.– Spectral graph methods for segmentation.– Appearance and
shape models.
Spectral graph methods for segmentation.
Component Analysis for CV & PR
– Fundamental matrix estimation and calibration.– Compression.–
Classification.– Dimensionality reduction and visualization.
• Signal Processing– Spectral estimation, system identification
(e.g. Kalman filter), sensor
array processing (e.g. cocktail problem, eco cancellation),
blind source
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 4
array processing (e.g. cocktail problem, eco cancellation),
blind source separation, …
• Computer Graphics– Compression (BRDF), synthesis,…
• Speech, bioinformatics, combinatorial problems.
-
2
• Computer Vision & Image Processing– Structure from
motion.– Spectral graph methods for segmentation.– Appearance and
shape models.Appearance and shape models
Component Analysis for CV & PR
– Fundamental matrix estimation and calibration.– Compression.–
Classification.– Dimensionality reduction and visualization.
• Signal Processing– Spectral estimation, system identification
(e.g. Kalman filter), sensor
array processing (e.g. cocktail problem, eco cancellation),
blind source
pp p
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 5
array processing (e.g. cocktail problem, eco cancellation),
blind source separation, …
• Computer Graphics– Compression (BRDF), synthesis,…
• Speech, bioinformatics, combinatorial problems.
• Computer Vision & Image Processing– Structure from
motion.– Spectral graph methods for segmentation.– Appearance and
shape models.
Component Analysis for CV & PR
– Fundamental matrix estimation and calibration.– Compression.–
Classification.– Dimensionality reduction and visualization.
• Signal Processing– Spectral estimation, system identification
(e.g. Kalman filter), sensor
array processing (e.g. cocktail problem, eco cancellation),
blind source
Dimensionality reduction and visualization
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 6
array processing (e.g. cocktail problem, eco cancellation),
blind source separation, …
• Computer Graphics– Compression (BRDF), synthesis,…
• Speech, bioinformatics, combinatorial problems.
• Computer Vision & Image Processing– Structure from
motion.– Spectral graph methods for segmentation.– Appearance and
shape models.
Component Analysis for CV & PR
– Fundamental matrix estimation and calibration.– Compression.–
Classification.– Dimensionality reduction and visualization.
• Signal Processing– Spectral estimation, system identification
(e.g. Kalman filter), sensor
array processing (e.g. cocktail problem, eco cancellation),
blind sourcecocktail problem
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 7
array processing (e.g. cocktail problem, eco cancellation),
blind source separation, …
• Computer Graphics– Compression (BRDF), synthesis,…
• Speech, bioinformatics, combinatorial problems.
cocktail problem
Independent Component Analysis (ICA)Sound
Source 1Mixture 1
Sound Source 2
Mixture 2
Output 1
Output 2ICA
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 8
Sound Source 3
Mixture 3
Output 3
-
3
• Computer Vision & Image Processing– Structure from
motion.– Spectral graph methods for segmentation.– Appearance and
shape models.
Component Analysis for CV & PR
– Fundamental matrix estimation and calibration.– Compression.–
Classification.– Dimensionality reduction and visualization.
• Signal Processing– Spectral estimation, system identification
(e.g. Kalman filter), sensor
array processing (e.g. cocktail problem, eco cancellation),
blind source
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 9
array processing (e.g. cocktail problem, eco cancellation),
blind source separation, …
• Computer Graphics– Compression (BRDF), synthesis,…
• Speech, bioinformatics, combinatorial problems.
Why CA for CV & PR?• Learn from high dimensional data and
few samples.
– Useful for dimensionality reduction.
(Everitt,1984)
• Easy to incorporate – Robustness to noise, missing data,
outliers (de la Torre & Black, 2003a)– Invariance to geometric
transformations (Frey et al. 99, de la Torre & Black,
2003b;, Cox et al. 2008)
– Non-linearities (Kernel methods) (Scholkopf & Smola,2002;
Shawe-Taylor & Cristianini,2004)
– Probabilistic (latent variable models)M lti f t i l (t ) (
& O’ &
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 10
features samples
• Efficient methods O( d n<
-
4
Generative Models
• Principal Component Analysis/Singular Value Decomposition
BCD
Decomposition1) Robust PCA/SVD, PCA with uncertainty and missing
data.2) Parameterized PCA3) Filtered Component Analysis4) Subspace
regression5) Kernel PCA
• K-means and spectral clustering
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 13
6) Aligned Cluster Analysis (ACA)• Non-Negative Matrix
Factorization• Independent Component Analysis.
Principal Component Analysis (PCA)(Pearson, 1901; Hotelling,
1933;Mardia et al., 1979; Jolliffe, 1986; Diamantaras, 1996)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 14
• PCA finds the directions of maximum variation of thedata based
on linear correlation.
• PCA decorrelates the original variables.
PCA
Tnn μ1BCdddD ...21
d=d=pixelspixels
nn= images= images
kdnd BD
kccc ......21
kbbb 21
1 dnk μC
•Assuming 0 mean data the basis B that preserve the maximum
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 15
Assuming 0 mean data, the basis B that preserve the
maximumvariation of the signal is given by the eigenvectors of
DDT.
BΛBDD Td d
Snap-shot Method & SVD• If d>>n (e.g. images 100*100
vs. 300 samples) no DDT.• DDT and DTD have the same eigenvalues
(energy) and
related eigenvectors (by D). • B is a linear combination of the
data! (Sirovich 1987)• B is a linear combination of the data!
• [α,L]=eig(DTD) B=D α(diag(diag(L))) -0.5ΛDαDDαDDDDαBBΛBDD
TTTT
TVUΣD
• SVD factorizes the data matrix D as:
BCD
TT UUΛDD
TT VVΛDD
(Beltrami, 1873; Schmidt, 1907; Golub & Loan, 1989)
(Sirovich, 1987)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 16
SVDPCA
diagonal
nnnkkd
T
ΣIVVIUUVΣU
VUΣD
TT
TT CCIBBCB
BCDnkkd
-
5
Error Function for PCA
(Eckardt & Young, 1936; Gabriel & Zamir, 1979; Baldi
& Hornik, 1989; Shum et al., 1995; de la Torre & Black,
2003a)
n
E BCDBdCB 2)(
• PCA minimizes the following function:
• Not unique solution:• To obtain same PCA solution R has to
satisfy:
TT CCIBB
CRCBRBˆˆˆˆ
ˆˆ 1
kk RBCCBRR 1
Fi
iiE BCDBcdCB, 1
21)(
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 17
• R is computed as a generalized k×k eigenvalue problem.
CCIBB
11 BRBRCC TT (de la Torre, 2006)
PCA/SVD in Computer Vision• PCA/SVD has been applied to:
– Recognition (eigenfaces:Turk & Pentland, 1991; Sirovich
& Kirby, 1987; Leonardis & Bischof, 2000; Gong et al.,
2000; McKenna et al., 1997a)
– Parameterized motion models (Yacoob & Black, 1999; Black
et al., 2000; Black, 1999; Black & Jepson, 1998)
– Appearance/shape models (Cootes & Taylor, 2001; Cootes et
al., 1998; Pentland t l 1994 J & P i 1998 C i & S l ff 1999
Bl k & J 1998 Bl &et al., 1994; Jones & Poggio, 1998;
Casia & Sclaroff, 1999; Black & Jepson, 1998; Blanz
&
Vetter, 1999; Cootes et al., 1995; McKenna et al., 1997; de la
Torre et al., 1998b; de la Torre et al., 1998b)
– Dynamic appearance models (Soatto et al., 2001; Rao, 1997;
Orriols & Binefa, 2001; Gong et al., 2000)
– Structure from Motion (Tomasi & Kanade, 1992; Bregler et
al., 2000; Sturm & Triggs, 1996; Brand, 2001)
– Illumination based reconstruction (Hayakawa, 1994)– Visual
servoing (Murase & Nayar, 1995; Murase & Nayar, 1994)–
Visual correspondence (Zhang et al., 1995; Jones & Malik,
1992)
C i i i
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 18
– Camera motion estimation (Hartley, 1992; Hartley &
Zisserman, 2000)– Image watermarking (Liu & Tan, 2000)– Signal
processing (Moonen & de Moor, 1995)– Neural approaches (Oja,
1982; Sanger, 1989; Xu, 1993)– Bilinear models (Tenenbaum &
Freeman, 2000; Marimont & Wandell, 1992)– Direct extensions
(Welling et al., 2003; Penev & Atick, 1996)
1-Robust PCA•Two types of outliers:
Sample outliers Intra-sample outliers(Xu & Yuille., 1995)
(de la Torre & Black, 2001b; Skocaj & Leonardis, 2003)
•Standard PCA solution (noisy data):
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 19
Robust PCA• Using robust statistics:
Pixel residual(de la Torre & Black, 2001b; de la Torre &
Black, 2003a)
n
i
d
pp
k
jjipjppirpca cbdE
1 1 1
),(),,( μCB
quadraticoutlieroutlier
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 20
meanBasis (B) &Coefficients(c)
robustrobust
-
6
Numerical Problems• No closed form solution in terms of an
eigen-equation.• Deflation approaches do not hold.
First eigenvector with
T
T11
uuAA
uuAA
222
1
'''
'
First eigenvector with
highest eigenvalue.
Second eigenvector with highest eigenvalue.
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 21
• In the robust case all the basis have to be computed
simultaneously (including the mean).
How to Optimize it?
n
i
d
pp
k
jjipjppirpca cbdE
1 1 1
),(),,( μCB
C
HCC
BHBB
rpcac
nn
rpcab
nn
E
E
11
11
)(max
)(max
2
2
Tii
rpca
Tii
rpca
Ediag
Ediag
ccH
bbH
c
b
• Normalized Gradient descent
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 22
(Blake & Zisserman, 1987)• Deterministic annealing methods
to avoid local minima.
Example
Statistical outlier
• Small region• Short amount of time
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 23
Robust PCA
Original PCA RPCA Outliers
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 24
-
7
Structure from Motion
More work on Robust PCA
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 25
• Robust estimation of coefficients (Black & Jepson, 1998;
Leonardis & Bischof, 2000; Ke & Kanade, 2004)
• Robust estimation of basis and coefficients (Gabriel &
Odoro, 1984; Croux & Filzmoser., 1981; Skocaj et al.,
2002;Skocaj & Leonardis, 2003; de la Torre & Black, 2001b;
de la Torre & Black, 2003a)
• Other Robust PCA techniques (sample outliers) (Campbell, 1980;
Ruymagaart, 1981; Xu & Yuille., 1995)
More work on Robust PCA
1- PCA with Uncertainty and Missing Data
• If weights are separable closed-form solutionTwwW
d
i
n
j
k
ssjisijijF
cbdwE1 1
2
12 )()()( BCDWCB, • Adding uncertainty
If weights are separable closed form solution.
productHadamard
wij
nd
0 W
D
n
n
dd
dd
221
111
r
r
r w
w
...2
1
w
cnccc www 21w ……
r cwwW
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 26
• Generalized SVD
productHadamard
dnd dd 1
rdw
(Greenacre, 1984; Irani & Anandan, 2000;)
General Case• For arbitrary weights no closed-form solution.
dpTppTpTp
n
iiii
TiiF
diag
diagE1
2
))(()(
))(()()()(
bCdwbCd
BcdwBcdBCDWCB,
(Wiberg, 1976 , Torre & Black, 2003a)
• Alternated least squares algorithms– Slow convergence, easy
implementation.
• Damped Newton Algorithm– Fast convergence.
B
CBBCDWCB,
12
212
][)(
||||||||)()(
EEvec
E FFF
p 1
I
repeat
EErepeat
)(10
1
22
22
gHxy
vg
vH
(Buchanan & Fitzgibbon., 2005)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 27
– H definite positive:
vvvv
CB
v
22
22
)1( ][)()( EE
vecvec nn
Iv
H
22
2E econvergencuntil
FFuntilI
10;
)()()(
yx
xygHxy
Related work
• Iterative (Wiberg, 1976; Shum et al., 1995; Morris &
Kanade, 1998; Aans et al.,2002; Guerreiro & Aguiar, 2002)
• Closed-form (Aguiar & Moura, 1999; Irani & Anandan,
2000)P f t i ti• Power factorization (Hartley & Schaalitzky,
2003)
• Bayesian estimation (L.Torresani & Bregler, 2004)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 28
-
8
• Learn a subspace invariant to geometric transformations?
2- Parameterized Component Analysis (PaCA) (de la Torre &
Black, 2003b)
. . .
• Data has to be geometrically normalized– Tedious manual
cropping.
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 29
Tedious manual cropping.
– Inaccuracies due to matching ambiguities.
– Hard to achieve sub-pixel accuracy.
Error function for PaCA
)()()(),,( 211
2
1caBc)af(x,daCB
WtppE
T
ttt
Basis ((BB) &) &Motion Regularization
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 30
coefficients ((cc))(warping)Regularization
2
3121 1
2
211 WWaΓacΓc
tat
T
t
L
ltct
EigenEye Learning
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 31
More examples• UPS dataset.
Random selection of 100 images (16×16 pixels). Incrementally
update until preserve 80% of the energy.
PaK PCAOriginal CongealingPaK-PCAOriginal Congealing
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
-
9
Improving facial landmark labeling•Hand label (red dots),
PaK-PCA label (yellow)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
More on Parameterized CA• Probabilistic model
– Search scales exponentially with the number of motion
parameters (Frey & Jojic, 1999a; Frey & Jojic, 1999b;
Williams & Titsias, 2004)
• Other continuous approaches.
• Invariant clustering
• Non-rigid motion
(Schewitzer, 1999; Rao, 1999; Shashua et al., 2002, Cox et al.
2008)
(Fitzgibbon & Zisserman, 2003)
(Baker et al., 2004)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 34
• Invariant recognition
• Invariant support vector machines• Parameterized Kernel
Component Analysis (De la Torre, 2008)
(Black & Jepson, 1998)
(Avidan, 2001)
3- Filtered Component Analysis(de la Torre et al.,2007b)
1) No local minimum in the expected place.
2) Many local minima
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 35
2) Many local minima.
Multi-band representation
• Texture classification (Nunes et al. ‘03, Freeman, Zalesny
& Van Gool, Leung & Malik ‘01, Cula & Dana ‘01, Varma
& Zisserman ‘02, De Bonet ’97, Heeger & Bergen ’95,
Portilla & Simoncelli ’00, Zhu et al. ‘98)
• Face recognition (Wang et al ’03 Hie et al ’04 Wiskott et al
’97 Lades et alFace recognition (Wang et al. 03, Hie et al. 04,
Wiskott et al. 97, Lades et al. ’93, Wechler et al. ’02, Zhao et
al. ‘98)
• Filters (Gabor, Wavelets, Volterra, Fourier transform,
…)Convolution
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 36
-
10
Multi-band representation
1) Global minimum in the
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 37
1) Global minimum in the expected place.
2) Distance between global and other minima is larger.
Filtered Component Analysis (FCA)
22
11
22
11 ||)(||||)(||),...,(
2
fbackgroundj
n
j
F
ff
n
iFE FμdFμdFF
Filters
Images
ConvolutionConvolution
jivecvecvecvec
jTi
iTi
0)()(1)()(
FFFF
No overlap between filters
No trivial solution (0)
F n
T 2
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 38
f i
TfPCAE
1
22
1
||)(|| μdF
Robustness of FCA Training: 100 images Testing: 120 images
Correct global minimum
Gray FCA (4) Gabor(4)
41 % 74 % 62%
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 39
Correct global minimum 41 % 74 % 62%14.59 26 19.683.28 1.4
1.92
Correct to 2nd minimum distance
Average number of local minima
Other work
• Incremental PCA (de la Torre et al., 1998b; Ross et al., 2004;
Brand, 2002; Skocaj & Leonardis, 2003; Champagne & Liu.,
1998; A. Levy, 2000)
Mixture of subspaces• Mixture of subspaces (Vidal et al., 2003;
Leonardis et al., 2002)• Changing the margin in SVM (Ashraf and
Lucey 2010)• Exponential family PCA (Collins et al. 01)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 40
-
11
4- Subspace Regression: From a Single Image to a Subspace
• Traditional subspace methods
• Subspace Regression (Kim et al. 2010)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 41
4- Subspace Regression (II)
frontal(s=0) Subject Subspace
subj=1
subj=2
subj=i
… … … … ……
… … … … ……
……
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 42
TestImage(s=0)
?Predict a subspace from a single image
Subspace Regression (II)
b1 b2 b3 b4 b5
• Generated samples for each pose
Optimi ation problem
1 2 3 4 5
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
• Optimization problem
Experiment I
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
ErrorMeasure
Baseline I(img -> img)
Baseline II(img -> subsp)
SubspaceRegression
Matlab®’ssubspace()
1.3507(1.2312)
1.4088(1.1645)
1.0860(1.0651)
-
12
Experiment II
• Predicting a Subspace for Illumination– CMU PIE data set– 60
aligned subjects– 19 different illuminations
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
Subspace tracking
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-201146
(Template Matching: 42.99)
IVT-SS: 38.41
Subspace Regression: 37.98
5-Kernel PCA
),,(),,(),( 32122
212121 zzzxxxxxx
• The kernel defines an implicit mapping (usually high
dimensional andnon-linear) from input to feature space so the data
becomes linearly
Feature spaceInput space
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 47
non linear) from input to feature space, so the data becomes
linearlyseparable.
• Computation in the feature space can be costly because it
is(usually) high dimensional– The feature space is typically
infinite-dimensional!
Kernel Methods• Suppose (.) is given as follows
• An inner product in the feature space is
• So, if we define the kernel function as follows, there is no
need to carry out (.) explicitly
• This use of kernel function to avoid carrying out ( )
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 48
• This use of kernel function to avoid carrying out (.)
explicitly is known as the kernel trick. In any linear algorithm
that can be expressed by inner products can be made nonlinear by
going to the feature space
-
13
Kernel PCA(Scholkopf et al., 1998)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 49
Generative Models
• Principal Component Analysis/Singular Value Decomposition
BCD
Decomposition1) Robust PCA/SVD, PCA with uncertainty and missing
data.2) Parameterized PCA3) Filtered Component Analysis4) Subspace
regression5) Kernel PCA
• K-means and spectral clustering
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 50
6) Aligned Cluster Analysis (ACA)• Non-Negative Matrix
Factorization• Independent Component Analysis.
The Clustering Problem• Partition the data set in c-disjoint
“clusters” of data points.
• Number of possible partitions
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
12
1
10421
)1(1),(
cn
iic
ccnS n
c
i
c
• NP-hard and approximate algorithms (k-means, hierarchical
clustering, mog, …)
K-means
FTE ||)(||),(0 MGDGM
(Ding et al., ‘02, Torre et al ‘06)
xyD
TMG xy
57
y
TG
yD
M xy
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
1
2
3
45
6
7
8
9
10
x
-
14
Spectral ClusteringAffinity Matrix
(Dhillon et al., ‘04, Zass & Shashua, 2005; Ding et al.,
2005, De la Torre et al ‘06)
FcTE ||)(||),(0 WMCΓCM
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 53
Fc0
)(DΓ )](...)()([ 21 ndddΓ Normalized Cuts (Shi & Malik
’00)Ratio-cuts(Hagen & Kahng ’02)
6- Aligned Cluster Analysis (ACA)• Mining facial expression
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
• Mining facial expression for one subject• Mining facial
expression for one subject
Problem
• Summarization
• Visualization
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
• Indexing
• Mining facial expression for one subjectLooking up Sleeping
SmilingLooking forwardWaking up
Problem
• Summarization
• Visualization
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
• Indexing
-
15
• Mining facial expression of one subject
Problem
• Summarization
• Embedding
I d i
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
• Indexing
• Mining facial expression for one subject
Problem
• Summarization
• Embedding
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
• Indexing
k-means and kernel k-means2||||),( FJ MGXGM
(MacQueen 67, Ding et al. 02, Dhillon et al. 04, Zass and
Shashua 05, De la Torre 06)
xyX
)(G
MG xy
24
57
y
G )))((()( 1n GGGGIKG TTtrJM xy
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
13
4 6
8
9
10
x)()( XXK T
Problem formulation for ACA (I)
)..[ 21 ssX )..[ 43 ssX )..[ 1 mm ss X
1 2 3 1Labels (G)
1s 2s 3s 4sStart and end of the segments (s)
s 2)(),,( FacaJ MGXGM )..[)..[)..[ 13221 ,...,, mm ssssss
XXX
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
-
16
Problem formulation for ACA (II)
2
)..[)..[)..[ ),...,,(),,( 13221 Fssssssaca mmJ MGXXXSGM
k
ccSS
m
ici mg ii
1
2
2)..[1
1X
Dynamic Time Alignment Kernel (Shimodaira et al. 01)
X[Si , Si+1) mc
X [Si , Si+1)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
[ i , i+1)
mc
Matrix formulation for ACA
GGGGILKL 1n )(with)( TTkmk trJ
)()( XXK T
men
ts
GHGGGHILWLK 1n )(with))o(( TTTaca trJ
ers
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
samples
segm
2371,0 H
2323RW
clus
te
segments 731,0 G
Optimizing ACA (forward step)• Efficient Dynamic Programming
i =23 i =25 i =29
2.11.81.7
2.41.21.8
2.41.91.5
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
maxw
Optimizing ACA (backward step)
)( max2wnO
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
-
17
Honey bee dance data (Oh et al. 08)
Three behaviors: 1‐waggle, 2‐left turn, 3‐right turn
Seq 1 Seq 2 Seq 3 Seq 4 Seq 5 Seq 6ACA 0.845 0.925 0.600 0.922
0.878 0.928
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
PS- SLDS (Oh et al 08) 0.759 0.924 0.831 0.934 0.904 0.910
HDP- VAR(1)-HMM (Fox et al 08)
0.465 0.441 0.456 0.832 0.932 0.887
Spectral Clustering 0.698 0.631 0.509 0.671 0.577 0.649
Facial image features• Active Appearance Models (Baker and
Matthews ‘04)
Appearance
Upper face
Shape• Image features
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
Lower face
• Cohn-Kanade: 30 people and five different expressions
(surprise, joy, sadness, fear, anger)
Facial event discovery across subjects
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
• Cohn-Kanade: 30 people and five different expressions
(surprise, joy, sadness, fear, anger)
Facial event discovery across subjects
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
ACA Spectral Clustering
(SC)0.87(.05) 0.56(.04)
• 10 sets of 30 people
-
18
Unsupervised facial event discovery
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
Clustering human motion
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
clustering of human motion II
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
Generative Models
• Principal Component Analysis/Singular Value Decomposition
BCD
Decomposition1) Robust PCA/SVD, PCA with uncertainty and missing
data.2) Parameterized PCA3) Filtered Component Analysis4) Kernel
PCA
• K-means and spectral clustering5) Aligned Cluster Analysis
(ACA)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 72
• Non-Negative Matrix Factorization• Independent Component
Analysis.
-
19
“Intercorrelations among variables are the bane of the
multivariate researcher’s struggle for meaning”
Cooley and Lohnes, 1971
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 73
Part-based Representation
The firing rates of neurons are never negative. Independent
representations.
NMF & ICA
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 74
NMF & ICA
Non-negative Matrix Factorization• Positive factorization.
• Leads to part-based representation.0||||)( CB,BCDCB, FE
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 75
Nonnegative Factorization
ij ijijdF
2
0,0)(min BC
CB Inference:
(Lee & Seung, 1999;Lee & Seung, 2000)
j
ij
ijijij )(
)(BVBDB
CC TT
Learning:
Tij
T
ijij )()(
BCCDC
BB
Derivatives:
ijijij
F )()( CBBCBC
TT
TTF )()( DCBCC
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
• Multiplicative algorithm can be interpreted as diagonally
rescaled gradient descent.
ijTjj )(BCCijij
ij
)()( DCBCCB
-
20
Generative Models
• Principal Component Analysis/Singular Value Decomposition
BCD
Decomposition1) Robust PCA/SVD, PCA with uncertainty and missing
data.2) Parameterized PCA3) Filtered Component Analysis4) Kernel
PCA
• K-means and spectral clustering5) Aligned Cluster Analysis
(ACA)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 77
• Non-Negative Matrix Factorization• Independent Component
Analysis.
Independent Component Analysis
• We need more than second order statistics to represent the
signal.
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 78
ICA
• Look for si that are independent.• PCA finds uncorrelated
variables, the independent
components have non Gaussian distributions
1 BWWDSCBCD(Hyvrinen et al., 2001)
components have non Gaussian distributions.• Uncorrelated
E(sisj)= E(si)E(sj)• Independent E(g(si)f(sj))= E(g(si))E(f(sj))
for any non-
linear f,g
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 79
PCA ICA
ICA vs PCA
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 80
-
21
Many optimization criteria
• Minimize high order moments: e.g. kurtosiskurt(W) = E{s4}
-3(E{s2}) 2
• Many other information criteria.
n
ii
n
iii S
11)(cBcd
Sparseness (e.g. S=| |)
(Olhausen & Field, 1996)• Also an error function:
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 81
(Chennubhotla & Jepson, 2001b; Zou et al., 2005; dAspremont
et al., 2004;)
• Other sparse PCA.
Basis of natural images
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 82
Denoising
Originalimage Noisy Image
(30% i )(30% noise)
Denoise(Wi filt ) ICA
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 83
(Wiener filter) ICA
Outline• Introduction• Generative models
– Principal Component Analysis (PCA) and extensions– K-means,
spectral clustering and extensions– Non-negative Matrix
Factorization (NMF)– Independent Component Analysis (ICA)
• Discriminative models– Linear Discriminant Analysis (LDA) and
extensions– Oriented Component Analysis (OCA)– Canonical
Correlation Analysis (CCA) and extensions
• A unifying view of CA
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 84
• A unifying view of CA• Standard extensions of linear
models
– Latent variable models.– Tensor factorization
-
22
Discriminative Models
• Linear Discriminant Analysis (LDA)7) Discriminative Cluster
Analysis ) y8) Multimodal Oriented Discriminant Analysis
• Oriented Component Analysis (OCA)• Canonical Correlation
Analysis (CCA)
9) Dynamical Coupled Component Analysis10) Canonical Time
Warping
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 85
Linear Discriminant Analysis (LDA)
C C
(Fisher, 1938;Mardia et al., 1979; Bishop, 1995)
BΛSBSBSBBSBB bb t
tT
T
J ||||)(
C
i
C
j
Tjijib
1 1
))(( μμμμS
n
i
Tii
Tt
1
ddDDS
c C
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 86
• Optimal linear dimensionality reduction if classes are
Gaussian with equal covariance matrix.
Tji
c
j
C
ijiw
i
)()(1 1
μdμdS
Error function for LDA
FTTT
LDAE ||)()(||),( 21
DBAGGGBA
(de la Torre & Kanade, 2006)
[d1 d2 ... dn]
Equations n×c Unknowns d×c
Ad=pixels
K=di
m
subs
pace
knnk
ijg
G
1G1
}1,0{
0...01...00...1
TG
c=cl
asse
s
n=samples
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
Equations n×c Unknowns d×c
• d>>n an UNDETERMINED system of equations!
(over-fitting)
15
20
6
8
7-Discriminative Cluster Analysis (DCA)(de la Torre &
Kanade, 2006)
FTTT
DCAE ||)()(||),,( 21
DBAGGGGBA
−10
−5
0
5
10 −10−5
05
10
−20
−15
−10
−5
0
5
10
15
YX
Z
−8 −6 −4 −2 0 2 4 6 8−10
−8
−6
−4
−2
0
2
4
6
X
Y
20PCA+k-means DCA
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 88
−10−5
05
10 −10−5
05
10−20
−15
−10
−5
0
5
10
15
20
Y
X
Z
−10
−5
0
5
10 −10−5
05
10
−20
−15
−10
−5
0
5
10
15
20
YX
Z
PCA+k means
-
23
Clustering faces20
40
60
80
100
TT GGGG 1)(
20 40 60 80 100 120 140
120
140
0
0.1
0.2
PCA
PCA DCA
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 89
−0.4−0.2
00.2
0.4
−0.2
0
0.2
0.4
0.6−0.2
−0.1
DCA vs. PCA+k-means
1
1.05
DCA
PCA+k−means
0.75
0.8
0.85
0.9
0.95
Acc
urac
y
PCA+k−means
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 90
5 10 15 20 25 30 35 400.65
0.7
0.75
Number of clusters (classes)
8- Multimodal Oriented Component Analysis (MODA)
• How to extend LDA to deal with:– Model class covariances.
(de la Torre & Kanade, 2005a)
– Multimodal classes.– Deal efficiently with huge covariance
matrices
(e.g. 100*100).
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 91
Multimodality
−5
0
5
10
−200−100
0100
200
−200
0
200−10
−5
MODA
10
20
30
40
2
4
6
8
10
LDA
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 92
0 10 20 30 40 50 60−40
−30
−20
−10
0
10
0 10 20 30 40 50 60−10
−8
−6
−4
−2
0
2
-
24
MODAB that MAXIMIZES the Kullback-Leibler divergence between
clusters among l
classes
Trj
ri
classesri
rj
rj
ri
ri
rj
ri
rj
Ttr 1111 )))())(((( 2121211212 BμμΣΣμμΣΣΣΣB
classes.
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 93
i
jiij Cr Cr
ijjiijiji j1 1 2
• 1 mode per class and equal covariances equivalent to LDA.
Optimization
1
1 ))()(()(i
iT
iTtrJ BABBΣBB
• Hard optimization problem
T(B)
)()()()(
00 BBBBB
JTJT
J(B)
• Iterative Majorization (Kiers, 1995; Leeuw, 1994)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 94
W1
W0
Related LDA work
• Face recognition (Belhumeur et al., 1997;Zhao, 2000;Martinez
& Kak, 2003)
• Small sample problem (Chen et al., 2000; Yu & Yang, 2001)•
Mixture (Hastie et al., 1995; Zhu & Martinez, 2006;)• Neural
approaches (Gallinari et al., 1991; Lowe & Webb, 1991)•
Heteroscedastic discriminant analysis (Kumar &
Andreou, 1998; Fukunaga, 1990; Mardia et al., 1979; Saon et al.,
2000;)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 95
Discriminative Models
• Linear Discriminant Analysis (LDA)7) Discriminative Cluster
Analysis ) y8) Multimodal Oriented Discriminant Analysis
• Oriented Component Analysis (OCA)• Canonical Correlation
Analysis (CCA)
9) Dynamical Coupled Component Analysis10) Canonical Time
Warping
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 96
-
25
Oriented Component Analysis (OCA)
T bΣbsignal
OCAb
• Generalized eigenvalue problem:
OCAT
OCAOCA
noiseOCA
signal
bΣb
bΣb
keki bΣbΣ
noise
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 97
Generalized eigenvalue problem:• boca is steered by the
distribution of noise
keki
(de la Torre et al., 2005a)
Representational Oriented Component Analysis (ROCA)
kT
kTk
ek
i
bΣbbΣb
1
1
jTj
jTj
e
i
bΣbbΣb
2
2
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 98
Discriminative Models
• Linear Discriminant Analysis (LDA)7) Discriminative Cluster
Analysis ) y8) Multimodal Oriented Discriminant Analysis
• Oriented Component Analysis (OCA)• Canonical Correlation
Analysis (CCA)
9) Dynamical Coupled Component Analysis10) Canonical Time
Warping
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 99
Canonical Correlation Analysis (CCA)
• Learn relations between multiple data sets? (e.g. find
features in one set related to another data set)
• Given two sets , CCA finds the pair of directions w and w that
maximize the correlation
ndnd and 21 YX
(Mardia et al., 1979; Borga 98)
of directions wx and wy that maximize the correlation between
the projections (assume zero mean data)
• Several ways of optimizing it:Ty
TTy
Tx
TTx
yTT
x
YwYwXwXw
YwXw
TT w0XXYX0
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 100
• An stationary point of r is the solution to CCA.
y
xddddT
ddddT w
ww
YY00XX
Β0YXYX0
A )()()()( 21212121 ,
ΒwAw
-
26
9- Dynamic Coupled Component Analysis (DCCA)
Data 1Data 1 Data 2Data 2
(de la Torre & Black, 2001a)
• Learning the coupling
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 101
Learning the coupling.• High dimensional data.• Limited training
data.
Solutions?• PCA independently and general mapping
PCA PCA
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 102
• Signals dependent signals with small energy can be lost.
DCCA
ˆ
n
iiicca
i
E1
2
1ˆˆˆ)ˆ,,,,ˆ,(
WcBμdμμCABB
ectio
nec
tion
BReconstructionReconstruction
B̂
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 103
n
iii
n
ii
Ti
i
ii 1
2
3121
2
2
1
)(WW
AccμdBc DynamicsDynamicsPro
jePr
oje
Dynamic Coupled Component Analysis
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 104
-
27
10- Canonical Time Warping (CTW)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
Canonical Correlation Analysis (CCA)(Hotelling 1936)
• CCA minimizes:different #rows, same #columns
TT
ndnd yx YX ,
CCASpatial transformation
2),(
F
Ty
TxyxccaJ YVXVVV b
yTT
y
xTT
xts IVYYV
VXXV
.
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
A least-square formulation for DTW
same #rows, different #columnsyx ndnd YX ,
2),(
F
Ty
TxyxdtwJ YWXWWW
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
Canonical Time Warping (CTW)
Reminder 2
2
),(
),(
F
Ty
Txyxdtw
F
Ty
Txyxcca
J
J
YWXWWW
YVXVVV
different #rows, different #columns
spatial transformation
F
Ty
Ty
Tx
TxyxyxctwJ YWVXWVVVWW ),,,(
2
yyxx ndnd YX ,
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
temporal alignment
by
Ty
Ty
Ty
xT
xTx
Txts I
VYWYWV
VXWXWV
..
-
28
Facial expression alignment
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
Facial expression alignment
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
Aligning human motion
Boxing O i bi t
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
Boxing Opening a cabinet
Aligning motion capture and video
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
-
29
Outline• Introduction• Generative models
– Principal Component Analysis (PCA) and extensions– K-means,
spectral clustering and extensions– Non-negative Matrix
Factorization (NMF)– Independent Component Analysis (ICA)
• Discriminative models– Linear Discriminant Analysis (LDA) and
extensions– Oriented Component Analysis (OCA)– Canonical
Correlation Analysis (CCA) and extensions
• A unifying view of CA
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 113
• A unifying view of CA• Standard extensions of linear
models
– Latent variable models.– Tensor factorization
The fundamental equation of CA
FTE ||)(||),(0 WBAWBA
Given two datasets : nxnd and XD
CC
FcrE ||)(||),(0 WBAWBA
Weightsfor rows
Weightsfor columns
Regressionmatrices XD )()(
C
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
AB
Properties of the cost function• E0(A,B) has a unique global
minimum (Baldi and Hornik-89).
• Closed form solutions for A and B are:
)()()( 22120 AWWWAAWAA TrTTTT ccctrE
))(()()( 221222120 BWWWWWBBWBB TTTTTtrE
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
))(()()(0 BWWWWWBBWBB rcccrrtrE
Principal Component Analysis (PCA)
• PCA finds the directionsof maximum variation ofthe data based
on linear
(Pearson, 1901; Hotelling, 1933;Mardia et al., 1979; Jolliffe,
1986; Diamantaras, 1996)
the data based on linearcorrelation.
• Kernel PCA finds the
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
directions of maximumvariation of the data inthe feature
space.
),,(),,(),( 32122
212121 zzzxxxxxx
-
30
PCA-Kernel PCA
FTE ||)(||),(0 WBAWBA
• Error function for KPCA: (Eckardt & Young, 1936; Gabriel
& Zamir, 1979; Baldi & Hornik, 1989; Shum et al., 1995; de
la Torre & Black, 2003a)
• The primal problem:
FcrE ||)(||),(0 WBAWBA F
TkpcaE ||)(||),( BADBA
)(D
)()()( 1 BDDBBBB TTTkpca trE
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
))()(()()( 1 ADDAAAA TTTkpca trE
)()()(kpca• The dual problem:
Linear Discriminant Analysis (LDA)(Fisher, 1938;Mardia et al.,
1979; Bishop, 1995)
C
i
C
j
Tjijib
1 1))(( μμμμS
n
BΛSBSBSBBSBB bb tTtTtrJ )()()( 1
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
• Optimal linear dimensionality reduction if classes are
Gaussian with equal covariance matrix.
n
i
Tii
Tt
1
ddDDS
Canonical Correlation Analysis (CCA)
• Given two sets , CCA finds the pair of directions wx and wy
that maximize the correlation between the projections (assume zero
mean data)
ndnd and 21 DX
(Fisher 36;Mardia et al., 1979;)
between the projections (assume zero mean data)
Td
Td
Tx
TTx
dTT
x
DwDwXwXwDwXw
T
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
FcT
rE ||)(||),(0 WBAWBA TTE ||)()(||)( 2
1
XBADDDBA
Canonical Correlation Analysis (CCA)
FTT
CCAE ||)()(||),( 2 XBADDDBA
0...01...00...1
TG
c=cl
asse
s
n=samples
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
• CCA is the same as LDA changing the label matrix by a new set
X
-
31
K-means
FcT
rE ||)(||),(0 WBADWBA (Ding et al., ‘02, Torre et al ‘06)
xyD
TBA xy
57
y
TA
yD
B xy
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
1
2
3
45
6
7
8
9
10
x
Normalized cuts
FcT
rE ||)(||),(0 WBAΓWBA
(Dhillon et al., ‘04, Zass & Shashua, 2005; Ding et al.,
2005, De la Torre et al ‘06)
)(DΓ )](...)()([ 21 ndddΓ
Affinity Matrix
Normalized Cuts (Shi & Malik ’00)Ratio-cuts(Hagen &
Kahng ’02)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
Other Connections
• The LS-KRRR (E0) is also the generative model for:– Laplacian
Eigenmaps, Locality Preserving projections, MDS,
Partial least-squaresPartial least squares, ….
• Benefits of LS framework:– Common framework to understand
difference and communalities
between different CA methods (e.g. KPCA, KLDA, KCCA, Ncuts)–
Better understanding of normalization factors and
generalizations– Efficient numerical optimization less than
θ(n3) or θ(d3), where n
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011
p ( ) ( )is number of samples and d dimensions
Outline• Introduction• Generative models
– Principal Component Analysis (PCA).– Non-negative Matrix
Factorization (NMF).– Independent Component Analysis (ICA).
• Discriminative models– Linear Discriminant Analysis (LDA).–
Oriented Component Analysis (OCA).– Canonical Correlation Analysis
(CCA).
• A unifying view of CASt d d t i f li d l
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 124
• Standard extensions of linear models– Latent variable models.–
Tensor factorization
-
32
Latent Variable Models
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 125
Factor Analysis• A Gaussian distribution on the coefficients and
noise is
added to PCA Factor Analysis.
k NpNp BcμdBcdI0,ccηBcμd
),|(),|()|()(
(Mardia et al., 1979)
• Inference (Roweis & Ghahramani, 1999;Tipping & Bishop,
1999a)
TT
d
k
E
diagNppp
BBμdμdd
0,cημ,
)))((()cov(
),...,,()|()(),|(),|()|()(
21
),|()( Vmcd|c Np
),( dcp Jointly Gaussian
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 126
11
1
)()()(
),|()(
BBIVμdBBBm
|
T
TT
p
PCA reconstruction low error.FA high reconstruction error (low
likelihood).
Ppca• If PPCA.• If is equivalent to PCA. TTTT BBBBBB 11
)()(0
dTE Iηη )(
0
• Probabilistic visual learning (Moghaddam & Pentland,
1997;)
2)(
2)(
1
21
2
21
21
2
)()()(21
21
2
)()(21
)2()2()2()2()()()(
2
1
211
kdk
ii
d
c
ddeeeedppp
k
i i
iTT
d
μdIBBμdμdμd T
ccc|dd
iT
i dBc
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 127
More on PPCA
• Tracking (Yang et al 1999; Yang et al 2000a; Lee et al 2005;
de la Torre et(Tipping & Bishop, 1999b; Black et al., 1998;
Jebara et al., 1998)
• Extension to mixtures of Ppca (mixture of subspaces).
Tracking (Yang et al., 1999; Yang et al., 2000a; Lee et al.,
2005; de la Torre et al., 2000b)
• Recognition/Detection (Moghaddam et al., 2000; Shakhnarovich
& Moghaddam, 2004; Everingham & Zisserman, 2006)
• PCA for the exponential family (collins et al., 2001)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 128
-
33
Tensor Factorization
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 129
Tensor faces(Vasilescu & Terzopoulos, 2002; Vasilescu &
Terzopoulos, 2003)
people
expressions
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 130
viewsilluminations
Eigenfaces• Facial images (identity change)
• Eigenfaces bases vectors capture the variability in facial
appearance (do not decouple pose, illumination, …)
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 131
Data Organization• Linear/PCA: Data Matrix
– Rpixels x images
– a matrix of image vectorsD
Pixe
ls
ImagesD
• Multilinear: Data Tensor– Rpeople x views x illums x express x
pixels
– N-dimensional matrix
Views
D
D
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 132
N dimensional matrix– 28 people, 45 images/person– 5 views, 3
illuminations,
3 expressions per personexilvpp ,,,iIl
lum
inat
ions
-
34
N-Mode SVD Algorithm
N = 3
pixelsxexpressxillums.xviews xpeoplex 51 UUUUU . ZD 432
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 133
PCA:
TensorFaces:
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 134
Strategic Data Compression = Perceptual Quality
• TensorFaces data reduction in illumination space primarily
degrades illumination effects (cast shadows, highlights)
• PCA has lower mean square error but higher perceptual
errorTensorFaces
Mean Sq. Err. = 409.153 illum + 11 people param.
33 basis vectors
PCA
Mean Sq. Err. = 85.7533 parameters
33 basis vectors
Original
176 basis vectors
TensorFaces
6 illum + 11 people param.66 basis vectors
• PCA has lower mean square error but higher perceptual
error
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 135
Acknowledgments• The content of some of the slides has been
taken from previous
presentations/papers of:– Ales Leonardis.– Horst BischofHorst
Bischof.– Michael Black.– Rene Vidal.– Anat Levin.– Aleix
Martinez.– Juha Karhunen.– Andrew Fitzgibbon.– Daniel Lee.– Chris
Ding.
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 136
– M. Alex Vasilescu.– Sam Roweis.– Daoqiang Zhang.– Ammon
Shashua.
-
35
CA
Thanks
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 137
Bibliography
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 138
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 139 Component Analysis for CV & PR F. De la Torre
CVPR Easter School-2011 140
-
36
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 141 Component Analysis for CV & PR F. De la Torre
CVPR Easter School-2011 142
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 143
Bibliography
Zhou F., De la Torre F. and Hodgins J. (2008) "Aligned Cluster
Analysis for Temporal Segmentation of Human Motion“ IEEE Conference
on Automatic Face and Gestures Recognition, September, 2008.
De la Torre, F. and Nguyen, M. (2008) “Parameterized Kernel
Principal Component Analysis: Theory and Applications to Supervised
and Unsupervised Image Alignment“ IEEE Conference on Computer
Vision and Pattern Recognition, June, 2008.
Component Analysis for CV & PR F. De la Torre CVPR Easter
School-2011 144