Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

31
Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger
  • date post

    20-Dec-2015
  • Category

    Documents

  • view

    214
  • download

    1

Transcript of Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Page 1: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Face Recognition with Harr Transforms and SVMs

EE645 Final Project

May 11, 2005

J Stautzenberger

Page 2: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Outline

i. Motivationii. Description of Face Recognition System

i. Overviewii. Feature Extraction iii. Haar Transform iv. SVM

iii. Experimentsi. Structureii. Data Setiii. Results

iv. Conclusions

Page 3: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Motivation

• Very active field in CS right now.• Applications in Security, Multimedia Retrieval, Human Computer

Interaction, …• Many “good” algorithms exist:

• Correlation – Nearest Neighbor• Eigenfaces – PCA based• Fisher Faces

• I am interested in doing real time face detection and feature extraction.

Page 4: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Proposed System Overview

• Proposed a system Using cascaded SVM using Haar wavelet features with feature selection done by Adaboost.

• Combination of simple to complex classifiers• Using SVMs for entire problem can lead to a lot useless

computations on easily distinguishable background patters.• Adaboost feature selection will cut out almost all background

patterns quickly.– “The aim of boosting is to improve the classification of any given simple

learning algorithm.” (Schapire)

– Training with Adaboost can be very slow if used for the complete classification algorithm though. – algorithm converges very slowly when examples become very hard.

Page 5: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Feature Extraction

• Two general types of feature selection:– Filter Methods - preprocessing steps performed independently of the

classification algorithm

– Wrapper Methods – search though feature space using criterion of the classification algorithm to select optimal features

• 2 popular filter methods:– Haar Transform - simple

– Gabor Transform – not simple

Page 6: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Haar Feature Extraction

• Haar Wavelet• Breaks down image into 4 sub-samples

– HH High passed in vertical and horizontal direction

– LH Low passed in vertical and high passed in horizontal

– HL High passed in vertical and low passed in horizontal

– LL Low passed in vertical and horizontal directions

LL LH

HL HH

LLLH

HL HH

LH

HL HH

Page 7: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Rectangle Features

• Rectangle Feature Examples– horizontal– vertical– diagonal

• The sum of the pixels which lie in the white rectangles are subtracted from the sum pixels in the grey rectangles

Page 8: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Haar Example

Page 9: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

SVMs

• The SVM determines the optimal hyperplane which maximizes the margin.

• The margin is the distance between the hyperplane and the nearest sample from the hyperplane.

• Decision function:

• α is the solutions from quadratic programming problem • Non-zero α is called a support vector

xxyxf i

L

i ii

Page 10: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Images

• Take some initial images for simple testing• Either create or Find a large Database

– 2100 images of 2 people– Yale Database B

• 5760 single light source images of 10 subjects

• under 576 viewing conditions (9 poses, shown in Figure (4), x 64 illumination conditions, shown in Figure (5)).

Page 11: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Illumination

Page 12: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Poses

Page 13: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Experiments

• Initial Test

• No Feature Selection

• 100 64x64 training images

• 3072 length feature vector

• 2000 test images

Page 14: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Training Results

Page 15: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Test Results

Page 16: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Yale Image Experiment 1

• Two subjects – 512 Training Images

– 512 Test Images

– No feature selection

– 3072 length feature vector

– Linear SVM

• Training– 18 support vectors

– No misclassified images

• Testing– All images classified correctly

Page 17: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Yale Image Experiment 2

• 10 test subjects• 1024 faces• 384 training faces• Only 2 subjects trained• 640 test faces• 10 subjects tested• Training

– 38 support vectors

– 0 misclassified

• Testing– All faces positively classified… very bad

Page 18: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Yale Experiment 3

• Same setup as before but this time with feature extraction• 1 level Haar transform

– 4 filtered images

– 3072 length feature vector

– No Feature Selection this time

• Training– 20 support Vectors

– None misclassified

• Testing– classification error rate was 0.020

– All positive labels classified correctly

Page 19: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Training

Page 20: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Testing

Page 21: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Yale Experiment 4

• Same setup as 3 but with 2 level Haar Transform• 2 level Haar transform

– 7 filtered images

– 3072 length feature vector

– No Feature selection this time

• Training– 36 support Vectors

– None misclassified

• Testing– classification error rate was 0.000

– Very good…

Page 22: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Training

Page 23: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Testing

Page 24: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Yale Experiment 5

• Same setup as 3 and 4 but now with feature selection• Feature Selection Algorithm

– 1 level Haar Transform

– Sum 4 filtered images

– 4 features

• Training – Nonlinear Support Vector (RBF)

– 372 support Vectors

– 9 misclassified

• Testing– classification error rate was 0.2391

– Positive examples badly mislabeled

Page 25: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Training

Page 26: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Testing

Page 27: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Yale Experiment 6

• Same setup as 5 but now with 16 selected features• Feature Selection Algorithm

– 4 level Haar Transform

– Sum 16 filtered images

– 16 features

• Training – Back to Linear Support Vector Machine

– 11 support Vectors

– 0 misclassified

• Testing– classification error rate was 0.1812

– Positive examples labeled correctly

Page 28: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Training

Page 29: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Testing

Page 30: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

Conclusions

• Feature Extraction is simple but very powerful

• Better feature selection for better error rate– Better rectangle filters– Use Boosting

• Eliminate background patterns• Reduce features

– Gabor Transform

• Better Testing Needed– Test against known results– Crop Images better

• Implement 3 layer system– Feature Extraction– Boosting (soft classier)– SVMs (hard classifier)

Page 31: Face Recognition with Harr Transforms and SVMs EE645 Final Project May 11, 2005 J Stautzenberger.

References

• [1] "Georghiades, A.S. and Belhumeur, P.N. and Kriegman, D.J.", "From Few to Many: Illumination Cone Models for Face Recognition under Variable Lighting and Pose", "IEEE Trans. Pattern Anal. Mach. Intelligence", 2001, 23, 6, "643-660".

• [2] Le, Duy Dinh, Satoh S., Feature Selection by AdaBoost for SVM-Based Face Detection.

• [3] F. Smeraldi, O. Carmona, J. Bigün. Saccadic Search with Gabor features applied to Eye Detection and Real-Time Head Tracking (1998).

• [4] P. Viola and M. Jones. "Fast and Robust Classification using Asymmetric AdaBoost and a Detector Cascade"