Title: Robust Statistics Why do we use the norms we do
1Robust StatisticsWhy do we use the norms we do?
- Henrik Aanæs
- IMM,DTU
- haa_at_imm.dtu.dk
A good general reference is Robust Statistics
Theory and Methods, by Maronna, Martin and Yohai.
Wiley Series in Probability and Statistics
TexPoint fonts used in EMF. Read the TexPoint
manual before you delete this box. AAAAA
2How Tall are You ?
3Idea of Robust Statistics
To fit or describe the bulk of a data set well
without being perturbed (influenced to much) by
a small portion of outliers. This should be done
without a pre-processing segmentation of the data.
Outliers can be interesting too!
4Line Example
5Robust Statistics in Computer VisionImage
Smoothing
Image by Frederico D'Almeida
6Robust Statistics in Computer VisionImage
Smoothing
7Robust Statistics in Computer Visionoptical flow
Play Sequence MIT BCS Perceptual Science Group.
Demo by John Y. A. Wang.
8Robust Statistics in Computer Visiontracking via
view geometry
Image 1
Image 2
9(No Transcript)
10(No Transcript)
11(No Transcript)
12Gaussian/ Normal DistributionThe Distribution We
Usually Use
- Nice Properties
- Central Limit Theorem.
- Induces two norm.
- Leads to linear computations.
- But
- Is fiercely influenced by outliers.
- Empirical distributions often have fatter
tails.
13Gaussians Just are Models Too
- Alternative title of this talk
14Error or ?-functionsConverting from Model-Data
Deviation to Objective Function.
15?-functions and MLA typical way of forming ?-
functions
16?-functions and ML IIA typical way of forming ?-
functions
17?-functions and ML IIA typical way of forming ?-
functions
18Typical ?-functionsWhere the Robustness in
Practice Comes From
- 2-norm
- 1-norm
- Huber norm
- Truncated quadratic
- Bi-Squared
General Idea Down weigh outliers, i.e. ?(x)
should be smaller for large x.
19Typical ?-functionsWhere the Robustness in
Practice Comes From
- 2-norm
- 1-norm
- Huber norm
- Truncated quadratic
- Bi-Squared
- Induced by Gaussian.
- Very non-robust.
- Standard distribution.
20Typical ?-functionsWhere the Robustness in
Practice Comes From
- 2-norm
- 1-norm
- Huber norm
- Truncated quadratic
- Bi-Squared
- Quite Robust.
- Convex.
- Corresponds to Median.
21The Median and the 1-Norm
22The Median and the 1-NormExample with 2
observations
23The Median and the 1-NormExample with 2
observations
24The Median and the 1-NormExample with more
observations
25Typical ?-functionsWhere the Robustness in
Practice Comes From
- 2-norm
- 1-norm
- Huber norm
- Truncated quadratic
- Bi-Squared
- Quite Robust.
- Convex.
- Corresponds to Median.
26Typical ?-functionsWhere the Robustness in
Practice Comes From
- 2-norm
- 1-norm
- Huber norm
- Truncated quadratic
- Bi-Squared
- Mixture of 1 and two norm.
- Convex.
- Has nice theoretical properties.
27Typical ?-functionsWhere the Robustness in
Practice Comes From
- 2-norm
- 1-norm
- Huber norm
- Truncated quadratic
- Bi-Squared
- Discards Outliers.
- For inliers works asGaussian.
- Has discontinues derivative.
28Typical ?-functionsWhere the Robustness in
Practice Comes From
- 2-norm
- 1-norm
- Huber norm
- Truncated quadratic
- Bi-Squared
- Discards Outliers.
- Smooth.
29Quantifying RobustnessA peak at tools for
analysis
Bias vs. Variance
30Quantifying RobustnessA peak at tools for
analysis
Related to variance, on the previous slide
31Quantifying RobustnessYou want to be robust over
a range of models
32Quantifying RobustnessA peak at tools for
analysis
- Other measures (Similar)
- Breakage Point How many outliers can an
estimator handle and still give reasonable
results. - Asymptotic bias What bias does an outlier impose.
33Back to Imageshere we have multiple models
To fit or describe the bulk of a data set well
without being perturbed (influenced to much) by
a small portion of outliers. This should be done
without a pre-processing segmentation of the data.
34Optimization Methods
- Typical Approach
- Find initial estimate.
- Use Non-linear optimization and/or EM-algorithm.
- NB In this course we have and will seen other
methods e.g. with guaranteed convergence
35Hough TransformOne off the oldest robust methods
in visionOften used for initial estimate.
Curse of Dimesionality PROBLEM
Example from MatLab help
36RanSaCSampling in Hough space, better for higher
dimensions
- RANdom SAmpling
- Consensus, RANSAC
- Iterate
- Draw minimal sample.
- Fit model.
- Evaluate model by Consensus.
- In a Hough setting
- 1. and 2. corresponds to finding a good bin in
Hough space. - 3. Corresponds to calculating the value.
Run RanDemo.m
37RansacHow many iterations
38Iteratively Reweighted Least Squares IRLSEM-type
or chicken and egg optimization