PCA Channel Student: Fangming JI u4082259 Supervisor: Professor Tom Geoden.

Slides:



Advertisements
Similar presentations
Face Recognition and Biometric Systems Eigenfaces (2)
Advertisements

Principal Component Analysis Based on L1-Norm Maximization Nojun Kwak IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008.
EigenFaces.
Face Recognition Ying Wu Electrical and Computer Engineering Northwestern University, Evanston, IL
Face Recognition CPSC UTC/CSE.
Face Recognition Method of OpenCV
What Makes a Patch Distinct?
As applied to face recognition.  Detection vs. Recognition.
Computer Vision Spring ,-685 Instructor: S. Narasimhan Wean 5403 T-R 3:00pm – 4:20pm Lecture #20.
Region labelling Giving a region a name. Image Processing and Computer Vision: 62 Introduction Region detection isolated regions Region description properties.
Pattern Recognition Topic 1: Principle Component Analysis Shapiro chap
Eigenfaces As we discussed last time, we can reduce the computation by dimension reduction using PCA –Suppose we have a set of N images and there are c.
Face Recognition using PCA (Eigenfaces) and LDA (Fisherfaces)
Project 4 out today –help session today –photo session today Project 2 winners Announcements.
Face Recognition Jeremy Wyatt.
Implementing a reliable neuro-classifier
Face Recognition Using Eigenfaces
Computer Vision I Instructor: Prof. Ko Nishino. Today How do we recognize objects in images?
Smart Traveller with Visual Translator for OCR and Face Recognition LYU0203 FYP.
Pattern Recognition. Introduction. Definitions.. Recognition process. Recognition process relates input signal to the stored concepts about the object.
Biometrics & Security Tutorial 6. 1 (a) Understand why use face (P7: 3-4) and face recognition system (P7: 5-10)
Face Recognition: An Introduction
Preprocessing Images for Facial Recognition Adam Schreiner ECE533.
Facial Recognition CSE 391 Kris Lord.
CS 485/685 Computer Vision Face Recognition Using Principal Components Analysis (PCA) M. Turk, A. Pentland, "Eigenfaces for Recognition", Journal of Cognitive.
Vision-Based Biometric Authentication System by Padraic o hIarnain Final Year Project Presentation.
Face Recognition Using Neural Networks Presented By: Hadis Mohseni Leila Taghavi Atefeh Mirsafian.
Eigenfaces for Recognition Student: Yikun Jiang Professor: Brendan Morris.
SVD(Singular Value Decomposition) and Its Applications
PCA & LDA for Face Recognition
Training Database Step 1 : In general approach of PCA, each image is divided into nxn blocks or pixels. Then all pixel values are taken into a single one.
Recognition Part II Ali Farhadi CSE 455.
Face Recognition and Feature Subspaces
Face Recognition and Feature Subspaces
Principles of Pattern Recognition
Multimodal Interaction Dr. Mike Spann
Access Control Via Face Recognition Progress Review.
1 Recognition by Appearance Appearance-based recognition is a competing paradigm to features and alignment. No features are extracted! Images are represented.
Using Support Vector Machines to Enhance the Performance of Bayesian Face Recognition IEEE Transaction on Information Forensics and Security Zhifeng Li,
Classification Course web page: vision.cis.udel.edu/~cv May 12, 2003  Lecture 33.
Face Recognition: An Introduction
1 Terrorists Face recognition of suspicious and (in most cases) evil homo-sapiens.
CSE 185 Introduction to Computer Vision Face Recognition.
A NOVEL METHOD FOR COLOR FACE RECOGNITION USING KNN CLASSIFIER
Supervisor: Nakhmani Arie Semester: Winter 2007 Target Recognition Harmatz Isca.
GENDER AND AGE RECOGNITION FOR VIDEO ANALYTICS SOLUTION PRESENTED BY: SUBHASH REDDY JOLAPURAM.
Irfan Ullah Department of Information and Communication Engineering Myongji university, Yongin, South Korea Copyright © solarlits.com.
An Improved Approach For Image Matching Using Principle Component Analysis(PCA An Improved Approach For Image Matching Using Principle Component Analysis(PCA.
2D-LDA: A statistical linear discriminant analysis for image matrix
Obama and Biden, McCain and Palin Face Recognition Using Eigenfaces Justin Li.
3D Face Recognition Using Range Images Literature Survey Joonsoo Lee 3/10/05.
Face detection and recognition Many slides adapted from K. Grauman and D. Lowe.
Part 3: Estimation of Parameters. Estimation of Parameters Most of the time, we have random samples but not the densities given. If the parametric form.
Septian Adi Wijaya – Informatics Brawijaya University
PRINCIPAL COMPONENT ANALYSIS (PCA)
Submitted by: Ala Berawi Sujod Makhlof Samah Hanani Supervisor:
Lecture 8:Eigenfaces and Shared Features
Face Recognition and Feature Subspaces
Recognition: Face Recognition
Principal Component Analysis
Outline Peter N. Belhumeur, Joao P. Hespanha, and David J. Kriegman, “Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection,”
In summary C1={skin} C2={~skin} Given x=[R,G,B], is it skin or ~skin?
Facial Recognition in Biometrics
Face Recognition and Detection Using Eigenfaces
PCA is “an orthogonal linear transformation that transfers the data to a new coordinate system such that the greatest variance by any projection of the.
Outline H. Murase, and S. K. Nayar, “Visual learning and recognition of 3-D objects from appearance,” International Journal of Computer Vision, vol. 14,
Feature space tansformation methods
Announcements Project 4 out today Project 2 winners help session today
Facial Recognition as a Pattern Recognition Problem
The “Margaret Thatcher Illusion”, by Peter Thompson
Presentation transcript:

PCA Channel Student: Fangming JI u Supervisor: Professor Tom Geoden

Organization of the Presentation  PCA and problems  PCA channel idea  Use the channel for automatic classification  Channel  Corrected Channel  Conclusion  Future work

Principle Component Analysis  A statistic tool  Maximizes the scatter of all projected samples in the image space.  Tries to capture the most important features and reduce the dimensions at the same time  Each eigenvector is a principle component

Algorithm of PCA  Given a training set of M images with the same size, convert each of them into a single dimension vector (I1, I2, … Im) Then, find the average image by calculating the mean of the training set Ψ = (∑In) / M, n = 1, …m. Each training image differs from the average by Φn = In - Ψ. Then, the covariance matrix C is found by  where A = [Φ1, Φ2, … Φm] and C is a matrix. It is too big to be used in practice. But fortunately, there are only M-1 non-zero eigenvalues and they can be found more efficiently with an M x M computation. This means that we can compute the eigenvector vi of instead of computing the eigenvector ui of. Also we can notice that the M best eigenvalues of are equal to the M best eigenvalues of. Then we can get M best eigenvalues of ui by Avi. At the end we will select a value K, to keep only K largest eigenvalues.

Eigenfaces

Problems of PCA based methods  Avalanche disaster  Up to a certain limit, these methods are robust over a wide range of parameter.  Algorithm breaks down dramatically beyond that point

Constant Features and Inconstant Features  Holistic features = Local features + inconstant features = Local features + inconstant features  Local Features (constant features)  Inconstant features (such as view, illumination and expressions)  Little change from inconstant => Little change for holistic one  Great change of inconstant => maybe great change for the holistic one

Distribution in the Image Space  Images from the same personality may sit in totally different regions of the images space.  Distance between the images beyond the range of being correctly recognized

The PCA Channel  Holistic features = Local features + inconstant features  Positions decided by both local features and inconstant features  Incremental changes in the inconstant features, should produce incremental changed holistic features or positions  This incremental changed position looks like a channel so we call it “PCA Channel”

Experiment Preparation And Tools  Collecting images with incremental changes in the orientations -- Mingtao’s software  45 images from three identities (15 images for each identity which are changed incrementally in orientation)  Dozens of images from another three identities, randomly oriented with some expression images  Face Recognition Practitioner – Software developed by me

Existence of The Channel  Take view for example

Automatic Image Classification  1)Given an input image  2)Recognize it  3) Compute the PCA again with the new recognized image  4) Go to step 1)  1) Give an input image  2) Recognize it  3) Put it into the training set  4) Go to step 1)  Original PCA method  The PCA channel method

Performance Comparison  If the training set is carefully selected the performance of PCA channel is better than the original one  Problems:  Sensitive to the selection of the training set  Contagious problem

Contagious Problem

The Corrected PCA Channel  Cut off the root of the mismatching  Improve the robustness

Implementation  Set up two threshold: Low(L) and High(H)  If the distance between the input image and the its nearest image in the training set H, put it for future recognition; if L H, put it for future recognition; if L < distance < H, make it a new group.  Calculate the PCA again and cut off the mismatching at here  Match again

Results  The success rate = Match to Original Training Set + Match to New Group  The success rate = 44.15%+50.65% = 94.80%  The success rate = 44.15%+51.94% = 96.09%  59.74%

New Groups

Conclusion  Properly build up image database and the PCA channel with cautious implementation, we can get very good performance for face recognition.  But from the above experiment we can see that, the strength but also the weakness of the PCA channel is the images database.  3D face reconstruction system.  Large computational load. But it can also be appropriate in some situations where the focus is more on accuracy than response time.

Future Works  Verify Our Research On Larger Data Set  Preprocess the images before recognition  Build Up a 3D-Face Morphable Model System  Research in Hybrid Methods