face recognition

23
FACE RECOGNITION Abdelfattah Al Zaqqa PSUT-2012

Upload: abdelfattah-al-zaqqa

Post on 07-Nov-2014

25 views

Category:

Documents


5 download

DESCRIPTION

Face Recognition and its techniques

TRANSCRIPT

Page 1: Face Recognition

FACE RECOGNITION

Abdelfattah Al Zaqqa PSUT-2012

Page 2: Face Recognition

Face Detection and Face Recognition

Face Detection, where a photo is searched to find any face (shown here as a green rectangle), then image processing cleans up the facial image for easier recognition.

Face Recognition, where that detected and processed face is compared to a database of known faces, to decide who that person is (shown here as red text). It uses algorithims.

Page 3: Face Recognition

Why Use the Face for Recognition

It works with the most obvious individual identifier – the human face.

its non-contact process since face images can be acquired from a distance by an inexpensive equipment; fixed camera. scan many people at once, no action required.

Other techniques need expensive equipments and too sensitive to body motion.

Other biometric methods require some voluntary action by the user i.e fixed position in front of a camera for iris identification.

Page 4: Face Recognition

Why Face for Recognition

Face recognition is totally non-intrusive and does not carry any such health risks.

Web has a huge amount of faces database.

techniques that rely on hands and fingers can be rendered useless if the epidermis tissue is damaged in some way (i.e., bruised or cracked).

Page 5: Face Recognition

How Face Recognition Works

Page 6: Face Recognition

How it Works

Detect and Capture - : we detect faces from image. Alignment and Extraction - Unique data is extracted

from the sample and a template is created Measurement: It measures the overall facial

structure, including distances between eyes, nose, mouth, and jaw edges.

Comparison - The template is then compared with a new sample

Matching - The system then decides if the features extracted from the new sample are matching or not

Identifying: Identify the

Page 7: Face Recognition

Face Recognition Techniques

Face recognition techniques can be roughly divided into two main categories: global approaches and feature based techniques. In global approaches the whole image serves as a feature vector, while in local feature approaches a number of fiducial or control points are extracted and used for classification.

Page 8: Face Recognition

1. global approaches

Global approaches model the variability of the face by analyzing its statistical properties based on a large set of training images. Representative global techniques are eigenfaces, Linear Discriminant Analysis (LDA), Support Vector Machines (SVM) and neural networks.

With the assistance of the eigenface algorithm, we can identify a person from a photo not in the database.

Page 9: Face Recognition

What is the distance between two images?

compare the images pixel by pixel would not work. For several reasons: Not each pixel holds valuable

information. For example background and hair pixels would arbitrary make the distance larger or smaller..

Also for direct comparison we would need the faces to be perfectly aligned in all pictures and we would hope that the rotation of the head was always the same. PCA algorithm overcome this issue.

Page 10: Face Recognition

PCA algorithm Principal component analysis

PCA algorithm creates a set of principal components, which are called eigenfaces.• Data is projected into a lower dimensional

space• preserving the directions that are most

significant• not necessarily orthogonal to the original

ones!

Eigenfaces are images, that represent the main differences between all the images in the database.

Page 11: Face Recognition

Eigenfaces

The recognizer first finds an average face by computing the average for each pixel in the image.

Each eigenface represents the differences from the average face. First eigenface will represent the most significant differences between all images and the average image and the last one the least significant differences.

Image1 = Avarage Image + 10% Eigenface 1 + 4% Eigenface 2 + … + 1% Eigenface 5

Page 12: Face Recognition

Eigenfaces

1. Compute the distance between the captured image and each of the images in the database.

2. Select the example from the database, which is closest to the processed image (the one with the smallest distance to captured image).

3. If the distance is not too big – label the image as concrete person.

Page 13: Face Recognition

Eigenfaces

The approach of using eigenfaces for recognition was developed by Sirovich and Kirby (1987) and used by Matthew Turk and Alex Pentland in face classification.

click here to see Example of FaceRecognitionUsingTheEigenfaceAlgorithm-source using Wolfram Mathimatica.(Wolfram Mathimatica must be install to execute the Example)

Page 14: Face Recognition

Explanition for Eigenface Algorithm

We have a face database of four people with photos of size 112×92 (10304 pixels). We represent each photo as a vector with 10304 elements to form a 10304×4 matrix . From each element of  we subtract the mean value of the elements from the same row to get a matrix . The four eigenvectors of the matrix  are calculated as are the coordinates of every "photo vector" in the basis formed by the eigenvectors.

Getting a set of 24 photos (six different photos of every person) we choose one of them, add some image effects if we want, and calculate its coordinates. It will correspond to one of the four photos we have with the minimum "distance" between them. So the given photo will be recognized or not.

Page 15: Face Recognition

Back propagation Neural Network (BPNN).

• simulating the behavior of human brain to perform machine learning tasks

• Basically Neural Network is a set of interconnected nodes.• The edges which are between the nodes are weighted so the

information which travels between two nodes is amplified.• The information travels from set of input nodes, across a set

of hidden nodes to a set of output nodes.• The intermediate (hidden-layer) neurons are as many as the

number of reduced dimensions that are intended.• The network “learns” what patterns are likely faces or not• Initially promising, but Cottrell and Fleming showed that they

can at best match an eigenface approach.

When BPNN technique is combined with PCA, non linear face images can be recognized easily. Hence it is concluded that this method has the acceptance ratio is more than 90 % and execution time of only few seconds.

Page 16: Face Recognition

Back propagation Neural Network (BPNN).

Page 17: Face Recognition

Neural Networks vs Neural Networks

Neural Networks are complex subject, but it has been shown that rarely they have better performance than eigenfaces algorithm. Sometimes the image is first defined as linear combination of eigenfaces and then it’s describing vector is passed to the Neural Network. In other words eigenfaces algorithm builds really the base of face recognition.

Page 18: Face Recognition

2- feature based techniques

The main idea behind feature-based techniques is to discriminate among different faces based on measurements of structural attributes of the face. Most recent approaches are the Embedded Hidden Markov Models (EHMMs), the Elastic Graph Matching and Dynamic Link Architecture.

Page 19: Face Recognition

General Difficulties it’s not always accurate

Recognize the face from a general view point under large variability of the recorded images due to illumination conditions, facial expressions, presence of glasses, aging effects etc.

Identify similar faces

Page 20: Face Recognition

Uses:

Easy people tagging: Facebook’s automatic tag suggestion feature.

Gaming: Microsoft’s Kinect’s it enables users to control and interact with the Xbox 360 without the need to touch a game controller.

Vending machines: vending machines suggest soft drinks based on stereotypes based on your gender and age (and the weather).

Billboards: billboards contain technology that figures out a person’s gender and age to within 10 years, and presents them with the appropriate advertising.

Security: Detecting fake ID cards, Searching mugshot databases

Places: airports (Frankfurt airport security system), banks, safes

Page 21: Face Recognition

Uses:

Revealing more about yourself in public.

Hotels and restaurants: technology companies, equip hotels and restaurants with the technology to let them flag VIP guests.

…etc.

Page 22: Face Recognition

Conclusions and Future Developments

The problem of machine face recognition has been an ongoing subject of research for more than 20 years.

The use of novel sensors, such as 3D, can help overcome limitations due to viewpoint and lighting variations. which can recognize you as soon as you walk in front of the camera.

On the other hand, it has been acknowledged that there is no perfect biometric and thus the combination of different modalities, e.g. face combined with speaker, fingerprint and/or hand recognition, is required to achieve the desired level of performance.

Page 23: Face Recognition

Feel free to ask any questions.

Questions?

Thanks