facial recognition vinod

21
Facial recognition Vinod V S (081006056)

Upload: vinod-siragaon

Post on 05-Dec-2014

2.056 views

Category:

Education


0 download

DESCRIPTION

 

TRANSCRIPT

Page 1: Facial Recognition Vinod

Facial recognition

Vinod V S (081006056)

Page 2: Facial Recognition Vinod

What is facial recognition ?

comparing

Result : matched or not

Image compared with the database

Page 3: Facial Recognition Vinod

Implementation

• Any characteristics is converted to binary templates and then it is stored in databases

• Later if verification is needed then these templates are compared

• There are many techniques to find templates

Characteristics

011001010010101…011010100100110…001100010010010...

Templates

Page 4: Facial Recognition Vinod

How does a human recognize people ?

• When we see somebody our eyes acts as a scanner and our brain as a database and we recognize them provided his image is in your brain

• When we see people's faces it activates a small region of the brain, known toneuro. When we look at any other type of object this tissue normally stays quiet.

• That means when we see a stranger’s face we go blank and cant recognize ,and the simple reason for that is that we don’t have his face in our database (BRAIN)

CAN YOU IDENTIFY HIM

Page 5: Facial Recognition Vinod

Facial Recognition• Every face has numerous, distinguishable landmarks, the different

peaks and valleys that make up facial features.

• FaceIt defines these landmarks as nodal points. Each human face has approximately 80 nodal points.

• Some of these measured by the software are • Distance between the eyes

• Width of the nose

• Depth of the eye sock

• The shape of the cheekbones

• The length of the jaw line

Page 6: Facial Recognition Vinod

Two problems in face recognition

In this photo we can visually see the

illumination problem ,the software

cannot recognize areas of dark illumination

In this photo we can see the person

is giving a pose which causes a

difficulty for the software

The pose problem

The illumination problem

Page 7: Facial Recognition Vinod

How to solve illumination problem ?

1. Heuristic methods including discarding the leading principal components

2. A photo is captured and it is normalized such that the pixels illumination gets increased and we get a clear picture for further processing .

Page 8: Facial Recognition Vinod

How to solve the pose problem ?

1. Multiple images based methods when multiple images per person are available

2. Here we can take a 2d photo and then reconstruct it in 3d and then we can changing the pose illumination and expression using software .This helps us to compare the same kind of pose or

expression

Page 9: Facial Recognition Vinod

3D facial recognition• Steps involved

1. DETECTION

2. ALIGNMENT

3. MEASUREMENT

4. REPRESENTATION

5. MATCHING

6. VERIFICATION

Page 10: Facial Recognition Vinod

Facial detection :DEMO

Page 11: Facial Recognition Vinod

Facial Detection• Faces decompose into 4 main

organs

• Eyebrows

• Eyes

• Nose

• Mouth

• Human skin has its own color distribution that differs from that of most of nonface objects.

Cameras also use the same technology to detect faces

Page 12: Facial Recognition Vinod

Alignment• Once it detects a face, the system determines the head's position, size

and pose.

• In real life we saw that we don’t get faces that are frontal and the software is not so intelligent that it can use those images and hence we align the expression ,pose as we need

Page 13: Facial Recognition Vinod

Measurement• After alignment is done, we now measure every detail of

the face we want to compare• The system measures the curves of the face on a sub-

millimeter (or microwave) scale and creates a template

Page 14: Facial Recognition Vinod

Representation• The system translates the template into a unique code.

This coding gives each template a set of numbers to represent the features on a subject's face.

• This unique code is what needed when we are going to compare the faces with the faces present in the database

CONVERTED

TEMPLATES

Page 15: Facial Recognition Vinod

Verification or Identification• In verification, an image is matched to only one image in

the database (1:1).

• If identification is the goal, then the image is compared to all images in the database resulting in a score for each potential match (1:N).

Page 16: Facial Recognition Vinod

Future Uses of Facial Recognition Systems• It will be mainly used in law enforcement agencies security

• Some government agencies have also been using the systems for security and to eliminate voter fraud

• The U.S. government has recently begun a program called US-VISIT (United States Visitor and Immigrant Status Indicator Technology), aimed at foreign travelers gaining entry to the United States.

Page 17: Facial Recognition Vinod

• Critics say it produces too many false positives

• Invasion of privacy

• To easy to misuse for wrong purposes

Not Everyone Loves Face Recognition

Page 18: Facial Recognition Vinod

Conclusion• At present it is most promising for small- or medium-scale

applications, such as office access control and computer log in; it still faces great technical challenges for large-scale deployments such as airport security and general surveillance

• Advancements in hardware and software needed

• Slow integration into society in limited environments

• Very large potential market

Page 19: Facial Recognition Vinod

References• J. Gilbert and W. Yang. A Real-Time Face Recognition System using

Custom VLSI Hardware. Harvard Undergraduate Honors Thesis in Computer Science, 1993.

• M. Turk and A. Pentland. Eigenfaces for Recognition. Journal of Cognitive Neuroscience, 3(1), 1991

• www.wickipedia.com

• www.howstuffworks.com

• www.facial-recognition.com

Page 20: Facial Recognition Vinod

Questions

Page 21: Facial Recognition Vinod