- Hardik Dhokai

( 07CE017 )

Guide: Prof. Ashish Patel

Computer Department
LDRP Institute of Technology and Research
Outline
Introduction
Goal of face detection
Challenges in face detection
Different Techniques used for face
Detection
Pros and Cons of each technique
Discussion and Future work
 Introduction
Face Detection
Identify and locate human faces in an image regardless
of their

Size
Expressions
Pose
Illumination
Orientation
and occlusion
where are the faces, if any?
 Cont...
Definition
The face detection problem can be defined as :
“Given as input an image, which could be a
digitized video signal or scanned photograph,
determine whether or not there are any human
faces in the image if there are, return their
location”.
Why Face Detection is Important?
First step for most automatic security system

First step in many surveillance systems

Face is the unique identity of a person

Example applications:

Automated security systems

Intelligence information
 Why Face Detection is Difficult?

Face Localization: A simplified detection problem with the

assumption that an input image contains only one face.

Facial feature detection: To detect the presence and

location of features such as eyes, nose, nostrils, eyebrow,
mouth, lips, ears, etc.

Face tracking: Continuously estimate the location and

possibly the orientation of face in an image sequence in real
time.

Facial expression recognition: concerns the affective

states of human’s like happy,sad,disgusted etc.
 Cont...

Size: small faces are more difficult to detect.

pose: frontal, 45 degree, profile, upside down.
Presence or absence of structural
beards, moustache, and glasses.

Orientation: Faces can appear in different orientations

within the image plane depending on the angle of
the camera and face.

occlusions: faces may be partially occluded by other

Some Examples of Partially Occluded Faces
Approaches to Detect faces

Knowledge-based Approach


Feature invariant Approach


Template matching Approach


Appearance-based Approach
 Knowledge-Based Approach

Represent a face using a set of human-coded
rules
Example:

The center part of face has uniform intensity values

A face often appears with two eyes that are
symmetric to each other, a nose and a mouth

Use these rules to guide the search process

Level 1 Level 2 Level 3

The input
Candidates of Basic face Face location&
image
faces locations segmentation
 Knowledge-based Approach:
Summary:

Pros:

Easy to come up with simple rules

Based on the coded rules, facial features in an input image
are extracted first, and face candidates are identified

Cons:

Difficult to translate human knowledge into rules

Difficult to extend this approach to detect faces in
different poses:
 Feature-Based Approach


Detect facial features (eyes, nose,
mouth, etc) first

Facial features: edge, intensity, shape,
texture, skin color, etc

Aim to detect invariant features

Group features into candidates and
verify them
 Feature-Based Approach :
Summary

Pros:

Features are invariant to pose and
orientation change


Cons:

Difficult to locate facial features due to
several corruption (illumination, occlusion)

Difficult to detect features in complex
background
 Template Matching Approach

Store a template
Predefined: based
on edges or regions

Deformable: based
on outline of face

Templates are hand-
coded (not learned)

Use correlation to locate faces
Template-Based Approach :
Summary

Pros:

It is simple to implement

Cons:

Templates needs to be initialized near
the face images

Difficult to enumerate templates for
different poses (similar to knowledge-
based methods)
 Appearance-Based Approach

Uses machine learning algorithms.

Fast and fairly robust techniques.
Train a classifier using positive examples of faces

Various Methods:
Neural model
Support vector machine (SVM)
Adaboost
Hidden Markov Model
Distribution Method
 CONCLUSION

- Face detection is still in a learning mode

- - We hope that in future we will have

robust Approaches/methods to detect

the face
- algorithms
 Web Resources

Face detection home page

http://www.facedetection.com/

http://vasc.ri.cmu.edu/NNFaceDetector/#upright

http://vision.ucsd.edu/kriegman-grp/research/ptrack/index.html

http://www.cvpr.org/2006/index.html

http://web.archive.org/web/20030602192115/http://www.cs.ru
References