Scribd
Upload
114 views

To Say .. A Simple Question It May Seem .: Why Use Face?

The document discusses face recognition techniques, including the advantages and disadvantages of 2D and 3D face recognition. 2D face recognition is sensitive to variations in illumination and facial expressions, while 3D techniques can overcome some of these challenges but require more advanced camera technology. Common algorithms for face recognition discussed include Eigenfaces and the Arena algorithm. Real-world applications of face recognition include passport authentication, criminal identification, and surveillance. The future of face recognition may include more advanced 3D techniques and wearable systems.

Uploaded by

Rutuja Satvase
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
114 views

To Say .. A Simple Question It May Seem .: Why Use Face?

The document discusses face recognition techniques, including the advantages and disadvantages of 2D and 3D face recognition. 2D face recognition is sensitive to variations in illumination and facial expressions, while 3D techniques can overcome some of these challenges but require more advanced camera technology. Common algorithms for face recognition discussed include Eigenfaces and the Arena algorithm. Real-world applications of face recognition include passport authentication, criminal identification, and surveillance. The future of face recognition may include more advanced 3D techniques and wearable systems.

Uploaded by

Rutuja Satvase
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 24

. To say .. a simple question it may seem.

Why use face?

Every one has got a fairly unique face Can be captured Without User Co operation.????
.....a question leading to the most efficient software.....

Features of 2-D and its challenges

Obstacles that must be overcome are differences in appearance due to variations in illumination, viewing angle, facial expressions, occlusion and changes over time.

Difference between 2 images of the same subject photographed from different angles is greater than the differences between 2 images of different subjects photographed from the same angle.
Recognition rates for unfamiliar faces drop significantly when there are different viewpoints for the training and test set .

From 2D to 3D face recognition


It is a much older research area than 3D face recognition research and broadly speaking. Information available in 3D face data means that 3D face recognition techniques might in the near future overtake 2D techniques.

Let us now look over 2-D and 3-D in detail

Disadvantages of 2-D face recognition


face recognition using 2D images is sensitive to illumination changes. The light collected from a face is a function of the geometry of the face, the albedo of the face, the properties of the light source and the properties of the camera. In 2D this task is nearly impossible due to the projective nature of 2D images. The databases used for 2-D recognition are in bulk and thus lead to lot of data mismatches at times even blunders. It is difficult to develop models that take all these variations into account.

Hence 3-D face recognition was introduced..

Advantages of 3-D face recognition.


Variations

in illumination only affect the texture of the face, yet the captured facial shape remains intact.
The physical dimensions of the face are known and are inherently encoded in the data. Better at capturing the surface geometry of the face.

Disadvantages of 3-D face recognition.

oily parts of the face with high reflectance may introduce artifacts under certain lighting on the surface. Requires cooperation from a subject. 3D sensor technology is currently not as mature as 2D sensors, the current computational cost of processing 3D data is higher than for 2D data.

Face Representation 2D vs. 3D (contd.)


2D Disadvantages >Cannot handle pose variation >Sensitive to lighting, variations shadows etc. 3D Disadvantages >Cameras are still expensive. >Takes time to reconstruct models >Unavailability of large collection of 3-D data

TECHNIQUES

Arena Algorithm

Eigen face

Procedure:
Skin Print: Picture Of Patch Of Skin Broken Into Smaller Blocks Differentiation On The Basis Of :- Any Lines :- Pores :-Actual Skin Texture Templates Used: Vector Template Local Feature Template Surface Texture Analysis

Advantages:
Help to find difference in identical twins increases accuracy by 20-25% Insensitive to changes in expression , blinking ,frowning , smiling Ability to compensate for mustache or beard growth & appearance of eye glasses Disadvantages: significant glare on the eye glasses Long hair obscuring the cetral part of face Poor lighting Lack of resolution

Arena Algorithm
Memory Based Technique For View Based Frontal Face Recognition Uses Principal Component Analysis And Neural Networks Closely Related To Correlation Templates Aim:- To Find Exact Match Of An Image From Given Datasets Reduced resolutions to 16x16, Reduces The Resolution Of 32x32 and 256x256 Image

Implementation
implementation of Arena algorithm implementation of neural networks depending upon outputs of Arena The programming language :- MATLAB Program for training :- arntrn.m Program for testing :- arntst.m

Eigenfaces.
The task of facial recognition is discriminating input signals
(image data) into several classes (persons) There are patterns in input signals such as presence of objects (eyes, nose, etc.,) and relative distance between these objects These characteristic features are EIGENFACES They can be extracted out of original image data by means of a mathematical tool called Principal Component Analysis (PCA)

The idea of principle component analysis to represent human faces was developed by Sirovich and Kirby (1987) and used by Mathew Turk and Alex Pentland for face detection & recognition Original face image can be reconstructed from eigenfaces if one adds up all the eigenfaces (features) in the right proportion In order to reconstruct the original image from the eigenfaces, one has to build a kind of weighted sum of all eigenfaces WEIGHT specifies to what degree a feature is present in the original face.. Use of weights to determine two important things *If the image in question is a face at all? *Similar faces can be grouped to clusters

Overview over the algorithm

Eigenfaces
:- The first successful example of facial recognition technology

:- Many extensions in the original method and many developments in automatic face recognition system :- Still often used as a baseline comparison method to demonstrate the minimum expected performance of such a system

REAL WORLD APPLICATIONS


PASSPORT AUTHENTICATION
CRIMINAL SCREENING. AVOID TRESSPASSING.

REAL WORLD APPLICATIONS


TO RECOGNISE THEIVES AND VIPS. PREVENTION OF VOTER FRAUD. SURVEILLANCE AT GOVERNMENT EVENTS. PICTURE PASSWORD . COMPANIES USE SURVEILLANCE CAMERAS FOR ADVERTISING STUDIES.

REAL WORLD APPLICATIONS


SONY CYBERSHOT CAMERA . WHICH CELEBRITY DO YOU RESEMBLE THE MOST ?

LIMITATIONS : scientist discover any protection software, whereas others find out more
excellent ways to avoid them test in Ybor city test at Boston Airport accuracy rate----61.4%.

Dependence on 2-D due to under development of 3-D effective and accurate only when image captured needed to be of a face that was looking almost directly at camera Face the Facts
Facial recognition technology? troubled past -- and troubling future.

Future of Face Recognition need of software for future generation. the ability to recognize a face at every transit point, such as air port passport control, train stations and metros, is key to our protection

Wearable Recognition system


wearable computing

software---recently tested by us army


Face Recognition for Smart Environments

machines---- functioning naturally.


the combination of 3D facial recognition and high speed database . for------any type of lightning.

future of google----------Neven vision. RIYA----replaced by Neven vision. some of the future facial recognition technique Paul Ekman---------facial action encoding system. John Ghent---------a professor at National university Ireland. company -----------Intelligent faces. plan of three package------ feature find, smart find, smart face pro.

You might also like