Face recognition process Face Recognition Biometric Systems 20052006
Face recognition process Face Recognition & Biometric Systems, 2005/2006
Plan of the lecture Face recognition process Most useful tools n n Principal Components Analysis Support Vector Machines Gabor Wavelets Hough Transform Biometric methods Face Recognition & Biometric Systems, 2005/2006
Face recognition process Detection Normalisation Feature vectors comparison Feature extraction Face Recognition & Biometric Systems, 2005/2006
Face detection: aims Find a face in the image n n n independent of image size independent of face size for RGB and GS images fast & effective independent from head rotation angle Face location passed to normalisation Face Recognition & Biometric Systems, 2005/2006
Face detection: tools Generalised Hough Transform n ellipse detection Support Vector Machines (SVM) n verification PCA (back projection) n verification Gabor Wavelets n feature points detection Colour-based face maps Face Recognition & Biometric Systems, 2005/2006
Face detection: algorithm Detection of ”vertical” ellipses n face candidates Detection of ”horizontal” ellipses n eye sockets candidates Initial normalisation and verification Detection of feature points Face Recognition & Biometric Systems, 2005/2006
Face tracking Useful in case of video sequences n n faster than detection smaller precision Tool: Optical flow Tracking of feature points Face Recognition & Biometric Systems, 2005/2006
Normalisation Input: n n image from a camera characteristic points location Target: n n generate an image of invariant parameters eliminate differences within classes Face Recognition & Biometric Systems, 2005/2006
Normalisation: tools Geometrical transforms Image filtering Histogram modifications n histogram fitting to a histogram of the average face image Lighting compensation Face Recognition & Biometric Systems, 2005/2006
Normalisation: stages Rotation of non-frontal faces Geometrical normalisation Lighting compensation Histogram fitting Face Recognition & Biometric Systems, 2005/2006
Feature extraction Input: n normalised image Target: n n generate a key which describes the face algorithm of comparing the keys Face Recognition & Biometric Systems, 2005/2006
Feature extraction: tools Principal Component Analysis n n n Linear Discriminant Analysis Local PCA Bayesian Matching Gabor Wavelets Face Recognition & Biometric Systems, 2005/2006
Feature vectors comparison Coherent with feature extraction Eigenfaces n n geometric distances SVM Dual Eigenfaces n image difference classified Elastic Bunch Graph Matching n correlation based Face Recognition & Biometric Systems, 2005/2006
Multi-method fusion Many feature extraction methods K 1 K 2. . . Kn Two images K 1 K 2. . . Kn Feature vectors S 1 S 2. . . Sn S Similarities Face Recognition & Biometric Systems, 2005/2006
Multi-method fusion Average similarity n weighted mean SVM with polynomial kernel SVM for finding optimal weights Face Recognition & Biometric Systems, 2005/2006
Tools: PCA Applications: n n n feature extraction – the Eigenfaces method detection (back projection) Dual Eigenfaces Stages: n n n training feature extraction feature vectors comparison Face Recognition & Biometric Systems, 2005/2006
Tools: SVM Applications: n n n face detection – verification feature vectors comparison detection of lighting direction estimation of head rotation angle multi-method fusion image quality assessment Face Recognition & Biometric Systems, 2005/2006
Tools: SVM Stages: n n training classification Main idea: n n data mapped into higher dimension to achieve linear separability mapping performed by application of kernels Problems with training set Parameters must be selected properly Face Recognition & Biometric Systems, 2005/2006
Tools: Gabor Wavelets Applications: n n n feature extraction (EBGM method) feature points detection face tracking (the detected points are tracked) Properties: n n n local frequency analysis set of various wavelets prepared comparison: correlation with displacement estimation Face Recognition & Biometric Systems, 2005/2006
Tools: GHT Useful for face detection Properties: n n n directional image generated (set of segments) probable ellipse centre for every segment (based on templates) accumulation of the results for all the segments in the image Face Recognition & Biometric Systems, 2005/2006
Biometric methods Types of the methods: n n static dynamic (behavioural) Requirements: n n n n universality distinctiveness permanence collectability performance acceptability circumvention Face Recognition & Biometric Systems, 2005/2006
Face recognition Advantages: n n n low invasiveness high speed identification support system Drawbacks: n n n relatively low effectiveness changeability of a face is not always visible Face Recognition & Biometric Systems, 2005/2006
Fingerprint recognition Advantages: n n high effectiveness useful forensic applications Disadvantages: n n long acquisition time low acceptability Face Recognition & Biometric Systems, 2005/2006
Iris recognition Advantages: n n high distinctiveness universality Drawbacks: n n high quality image required low permanence in young age Face Recognition & Biometric Systems, 2005/2006
Behavioural methods Gait recognition Voice recognition Signature analysis Face Recognition & Biometric Systems, 2005/2006
Thank you for your attention! Face Recognition & Biometric Systems, 2005/2006
- Slides: 26