Face Recognition and Biometric Systems
Transcript of Face Recognition and Biometric Systems
![Page 1: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/1.jpg)
Face normalisation
Face Recognition and Biometric Systems
![Page 2: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/2.jpg)
Plan of the lecturePlan of the lecture
Normalisation – task definitiontesting issuestesting issues
Geometric normalisationLighting normalisationAdvanced normalisation issuesAdvanced normalisation issues
Face Recognition and Biometric Systems
![Page 3: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/3.jpg)
Face recognition processFace recognition process
Detection Normalisation
FeatureFeature vectors Feature extraction
Feature vectorscomparison
Face Recognition and Biometric Systems
![Page 4: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/4.jpg)
Normalisation generalNormalisation – general
Image preparation for feature extraction
similar properties of generated imagesgeometryconditions (e.g.: lighting, expression)occlusions
Intra-class differences minimisedExtra-class differences notExtra class differences not influenced
Face Recognition and Biometric Systems
![Page 5: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/5.jpg)
Normalisation generalNormalisation – general
Effectiveness criteriavisual effectrecognition performanceg p
Detection error influences normalisation resultnormalisation result
Face Recognition and Biometric Systems
![Page 6: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/6.jpg)
Normalisation generalNormalisation – general
Perfect detectionreal location of face and facial featuresdata input by humanp y
Elimination of detection error propagationpropagationBetter assessment of subsequent recognition stages
Face Recognition and Biometric Systems
![Page 7: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/7.jpg)
Geometric normalisationGeometric normalisationR i tRequirements:
constant image sizefixed eye positionsfixed eye positionsfrontal orientation (soft requirement)
Frontal faces – goal:given positions of eyesaffine transform
Actions:Actions:clippingrotationscaling
Time for example
Face Recognition and Biometric Systems
![Page 8: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/8.jpg)
![Page 9: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/9.jpg)
![Page 10: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/10.jpg)
![Page 11: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/11.jpg)
![Page 12: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/12.jpg)
![Page 13: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/13.jpg)
![Page 14: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/14.jpg)
![Page 15: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/15.jpg)
Geometric normalisationGeometric normalisationS d ti i tiSpeed optimisation
larger image = more time consumed
Optimal algorithm:Calculate rotation angle1. Calculate rotation angle
2. Find and clip the ROI3. Rotate the clipped image4. Clip againp g5. Scale to the defined size
Face Recognition and Biometric Systems
![Page 16: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/16.jpg)
Laboratory reference (ex 2)Laboratory reference (ex 2)
Function parametersE itiEye positions:
left (49, 24)right (15, 24)
IPP referenceIPP referenceRotateCenterResizeResize
Operations...Face Recognition and Biometric Systems
![Page 17: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/17.jpg)
Lighting normalisationLighting normalisation
Lighting codnitions affect effectivenessNormalisation techniques:Normalisation techniques:
global filteringlocal modificationslocal modifications
Histogram modifications:stretchingstretchingequalisationfitting to the average face histogramfitting to the average face histogram
Filtering
Face Recognition and Biometric Systems
![Page 18: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/18.jpg)
Lighting normalisationLighting normalisation
Average faceAverage face
∑M
iiM 1
1= xμ=iM 1
M – number of faces in a setx – a single face vectorg
Face Recognition and Biometric Systems
![Page 19: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/19.jpg)
Lighting normalisationLighting normalisation
With histogram fitting:
Without histogram fitting:
Face Recognition and Biometric Systems
![Page 20: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/20.jpg)
Lighting normalisationLighting normalisation
Face Recognition and Biometric Systems
![Page 21: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/21.jpg)
Lighting normalisationLighting normalisation
Brightening filters – example of effects
Face Recognition and Biometric Systems
![Page 22: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/22.jpg)
Lighting normalisationLighting normalisation
Directional lighting:strong influence on the imagerecognition effectiveness much worseg
Light direction normalisation:li ht l d t tilight angle detectioncompensation to the frontal light conditions
Face Recognition and Biometric Systems
![Page 23: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/23.jpg)
Lighting normalisationLighting normalisation
Face Recognition and Biometric Systems
![Page 24: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/24.jpg)
Lighting normalisationLighting normalisation
Mirror reflectionLighting compensation masks:
lighting symmetrisationlighting symmetrisationcompensation to the average
d l b d kmodel-based mask
Filtering based on lighting modelg g gCompensation based on lighting model
Face Recognition and Biometric Systems
![Page 25: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/25.jpg)
Lighting normalisationLighting normalisation
Mirror reflectionCondition: no information in one image halfI h lfImage half recoveryApplicable to frontal faces onlyfaces onlyBrightness and angle thresholdingthresholding
Face Recognition and Biometric Systems
![Page 26: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/26.jpg)
Li hti li ti kLighting normalisation - masks
Image-based lighting compensation masksd k li ht ddark areas lightenedhighlights darkened
k h lMask imposition on the original image:additionmultiplicationadvanced imposition – to be investigated...
Face Recognition and Biometric Systems
![Page 27: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/27.jpg)
Li hti li ti k
S t i k
Lighting normalisation - masks
Symmetric mask
Compensation to the averageCompensation to the average
Face Recognition and Biometric Systems
![Page 28: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/28.jpg)
Lighting normalisationLighting normalisation
Lighting compensation – face modelDetection of lighting direction
based on average 3D face modelbased on average 3D face modelclassifiers (SVM, PCA)
C ti b d 3D d lCompensation based on 3D modelmask generationg
Works correctly for artificial data
Face Recognition and Biometric Systems
![Page 29: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/29.jpg)
Lighting normalisationLighting normalisation
Lighting-model based compensationLighting model based compensationInitial object (ambient light):
[ ]a[m,n]
Illuminated object:jc[m,n] = a[m,n] • I[m,n]
Face Recognition and Biometric Systems
![Page 30: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/30.jpg)
Lighting normalisationLighting normalisation
Light – low frequencies in the imageg q gLow frequencies elimination:
ln(c[m n]) ln(I[m n]) + ln(a[m n])ln(c[m,n]) = ln(I[m,n]) + ln(a[m,n])HP{ln(c[m,n])} ≈ ln(a[m,n])a’[m,n] = exp{HP{ln(c[m,n])}
Theory seems niceTheory seems nice...
Face Recognition and Biometric Systems
![Page 31: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/31.jpg)
Advanced normalisationAdvanced normalisation
Head rotation normalisationf t l i d i dfrontal image desired
Face expression normalisationneutral expressionexpression detectionp
Elimination of occlusionsglassesglassesbeard and moustache
Face Recognition and Biometric Systems
![Page 32: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/32.jpg)
Non frontal imagesNon-frontal images
Significant influence on recognition effectivenessNormalisation (rotation):
3D2D + depth map
The most serious problem: angle detectionFace Recognition and Biometric Systems
p g
![Page 33: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/33.jpg)
SummarySummary
Normalisation – important step in face itirecognition process
Tasks:as ssize and position normalisationimage properties normalisationimage properties normalisation
Many areas for further research
Face Recognition and Biometric Systems
![Page 34: Face Recognition and Biometric Systems](https://reader031.fdocuments.us/reader031/viewer/2022012016/61da926b674c2640f0738cd1/html5/thumbnails/34.jpg)
Thank you for your attention!Thank you for your attention!
Face Recognition and Biometric Systems