Title: Optimal Component Analysis Optimal Linear Representations of Images for Object Recognition
1Optimal Component AnalysisOptimal Linear
Representations of Images for Object Recognition
X. Liu, A. Srivastava, and Kyle Gallivan,
Optimal linear representations of images for
object recognition, IEEE Transactions on Pattern
Recognition and Machine Intelligence, vol. 26,
no. 5, pp. 662666, 2004.
2Outline
- Motivations
- Optimal Component Analysis
- Performance measure
- MCMC stochastic algorithm
- Experimental Results
- Fast Implementation through K-means
- Some applications
- Conclusion
3Motivations
- Linear representations are widely used in
appearance-based object recognition applications - Simple to implement and analyze
- Efficient to compute
- Effective for many applications
4Standard Linear Representations
- Principal Component Analysis
- Designed to minimize the reconstruction error on
the training set - Obtained by calculating eigenvectors of the
co-variance matrix - Fisher Discriminant Analysis
- Designed to maximize the separation between means
of each class - Obtained by solving a generalized eigen problem
- Independent Component Analysis
- Designed to maximize the statistical independence
among coefficients along different directions - Obtained by solving an optimization problem with
some object function such as mutual information,
negentropy, ....
5Standard Linear Representations - continued
- Standard linear representations are sub optimal
for recognition applications - Evidence in the literature 12
- A toy example
- Standard representations give the worst
recognition performance
6Proposed Approach
- Optimal Component Analysis (OCA)
- Derive a performance function that is related to
the recognition performance - Formulate the problem of finding optimal
representations as an optimization one on the
Grassmann manifold - Use MCMC stochastic gradient algorithm for
optimization
7Performance Measure
- It must have continuous directional derivatives
- It must be related to the recognition performance
- It can be computed efficiently
- Based on the nearest neighbor classifier
- However, it can be applied to other classifiers
as it forms clusters of images from the same
class that far from clusters from other classes - See an example for support vector machines
8Performance Measure - continued
- Suppose there are C classes to be recognized
- Each class has ktrain training images
- It has kcross cross validation images
9Performance Measure - continued
- h is a monotonically increasing and bounded
function - We used h(x) 1/(1exp(-2bx)
- Note that when b ? ?, F(U) is exactly the
recognition performance using the nearest
neighbor classifier - Some examples of F(U) along some directions
10Performance Measure - continued
- F(U) depends on the span of U but is invariant to
change of basis - In other words, F(U)F(UO) for any orthonormal
matrix O - The search space of F(U) is the set of all the
subspaces, which is known as the Grassmann
manifold - It is not a flat vector space and gradient flow
must take the underlying geometry of the manifold
into account see 3 4 5 for related work
11Deterministic Gradient Flow - continued
- Gradient at J (first d columns of n x n
identity matrix)
12Deterministic Gradient Flow - continued
- Gradient at U Compute Q such that QUJ
- Deterministic gradient flow on Grassmann manifold
13Stochastic Gradient and Updating Rules
- Stochastic gradient is obtained by adding a
stochastic component - Discrete updating rules
14MCMC Simulated Annealing Optimization Algorithm
- Let X(0) be any initial condition and t0
- Calculate the gradient matrix A(Xt)
- Generate d(n-d) independent realizations of wijs
- Compute Y (Xt1) according to the updating rules
- Compute F(Y) and F(Xt) and set dFF(Y)- F(Xt)
- Set Xt1 Y with probability minexp(dF/Dt),1
- Set Dt1 Dt / g and set tt1
- Go to step 1
15The Toy Example
- The following result on the toy example shows the
effectiveness of the algorithm - The following figure shows the recognition
performance of Xt and F(Xt)
16ORL Face Dataset
17Experimental Results on ORL Dataset
- Here the size of image is 92 x 112, d 5
(subspace) - Comparison using gradient, stochastic gradient,
and the proposed technique with different initial
conditions
18Results on ORL Dataset - continued
- With respect to d and ktrain
d20 ktrain5
d3 ktrain5
d10 ktrain5
d5 ktrain2
d5 ktrain8
d5 ktrain1
19Results on CMU PIE Dataset
- Here we used part of the CMU PIE dataset
- There are 66 subjects
- Each subject has 21 pictures under different
lighting conditions
20Some Comparative Results on ORL
- Comparison where performance on cross validation
images is maximized - In other words, the comparison is to show the
best performance linear representations can
achieve - PCA black dotted ICA red dash-dotted
- FDA green dashed OCA blue solid
21Some Comparative Results on ORL - continued
- Comparison where the performance on the training
is optimized - In other words, it is a fair comparison
- PCA black dotted ICA red dash-dotted
- FDA green dashed OCA blue solid
22Sparse Filters for Recognition
- The learning algorithm can be generalized to
other manifolds using a multi-flow technique
(Amit, 1991) - Here we use a generalized version to learn linear
filters that are sparse and effective for
recognition
23Sparse Filters for Recognition - continued
- Sparseness has been realized as an important
coding principle - However, our results show sparse filters are not
effective for recognition - Proposed technique
- To learn filters that are sparse and effective
for recognition
24Results for Sparse Filters
l1 1.0 and l2 -1.0
25Results for Sparse Filters - continued
l1 1.0 and l2 0.0
26Results for Sparse Filters - continued
l1 0.0 and l2 1.0
27Results for Sparse Filters - continued
l1 0.2 and l2 0.8
28Comparison of Commonly Used Linear Representations