A SEMINAR
ON
FACE RECOGNITION USING NEURAL NETWORK
Presented by-
Soumyajit Sarkar(Roll No-16900315100)
Tithi Dan(Roll No-16900315119)
Mentor :- Prof. SubhamPramanik
Department of Electronics & Communication Engineering
 Introduction
 History
 Face recognition
 Neural network
 TECHNOLOGICAL IDEAS
 ADVANTAGES
 DISADVANTAGES
 APPLICATIONS
 CONCLUSIONS
 REFERENCES
2
06/04/2018
 In the present scenario , there is great
need to maintain information security
or protection for physical property.
 When credit and ATM cards are lost or
stolen, an unauthorized user can often
come up with the correct personal
codes.
3
06/04/2018
 Face Recognition is the fastest verification technology as it works
with the most obvious individual i.e. The human face .
 Information and property can be secured through verification of
“true” individual identity.
 It consists of unique shape analysis, pattern and positioning of facial
features.
4
06/04/2018
 In 1960s, the first semi-automated system for facial recognition to locate
the features(such as eyes, ears, nose and mouth) on the photographs.
 In 1970s, Goldstein and Harmon used 21 specific subjective markers
such as hair colour and lip thickness to automate the recognition.
 In 1988, Kirby and Sirovich used standard linear algebra technique, to
the face recognition.
5
06/04/2018
 Face recognition involves comparing an image with a database of stored faces in order to
identify the individual in that input image.
6
 Face recognition involves comparing an image with a database of stored faces in order to
identify the individual in that input image.
06/04/2018
 Face recognition technology is the least intrusive. It works with the most obvious individual identifier-the
human face.
 It requires no physical interaction on behalf of the user.
 It can use your existing hardware infrastructure , existing cameras and image capture devices will work with
no problems.
7
06/04/2018
 It is a system of programs and data structures that approximates the operation of the
human brain.
8
06/04/2018
 Adaptive learning: An ability to learn how to do task.
 Self-Organization: Neural Network can create its own organisation.
 Remarkable ability to derive meaning from complicated or imprecise
data.
9
06/04/2018
 Here recognition is performed by both Principal Component Analysis (PCA) and
Back propagation Neural Networks(BPNN). All these processes are implemented for
Face Recognition, based on the basic block diagram as shown in fig 1.
BASIC BLOCK DIAGRAM(fig-1)
Pre-processed
input image
Back
Propagation
Neural
Network
(BPNN)
Principal
Component
Analysis
(PCA)
Classified
Output Image
10
06/04/2018
 It normalize and enhance the face image to improve the
recognition performance.
11
06/04/2018
 PCA is a common statistical technique for finding the patterns in high
dimensional data’s Feature extraction, also called Dimensionality Reduction.
12
06/04/2018
 Step 1: Partition face images into sub-patterns .
13
06/04/2018
 Step 2: Compute the expected contribution of each sub-pattern
 Generate the Mean and Median faces for each person, and use these
“virtual faces” as the probe set in training
 Use the raw face-image sub-patterns as the gallery set in for training,
and compute the PCA’s projection matrix on these gallery set
 For each sample in the probe set, compute its similarity to the samples
in corresponding gallery set
14
06/04/2018
 If a sample from a sub-pattern’s probe set is correctly classified, the
contribution of this sub-pattern is added by 1.
15
06/04/2018
When an unknown face image comes in -
 partition it into sub-patterns.
 classify the unknown sample’s identity
in each sub-pattern .
 Incorporate the expected contribution
and the classification result of all sub-patterns to generate the final
classification result.
16
06/04/2018
 It trains the network to achieve a balance between the ability to respond
correctly to the input patterns that are used for training & the ability to
provide good response to the input that are similar.
 It requires a dataset of
the desired output for
many input, making up
the training set.
17
06/04/2018
 These are necessarily Multilayer
Perceptrons (MLPs).
 MLPs:
1. Set of input layers
2. One or more hidden layers
3. Set of output layers
18
06/04/2018
 In a nutshell, face recognition is done in this way-
19
06/04/2018
 Fastest security mechanism.
 No physical interaction .
 more user friendly.
 no extra learning process.
 Simple, Fast & Easy to use.
 Social acceptability.
20
06/04/2018
 Identical twins attack
 Requires straight on, natural expression
 Affected by environment
21
06/04/2018
 criminals identification in public location
such as airport, Banks.
 Building security
 Credit card verification
 Mobile phone unlocking
22
06/04/2018
 We have to improve accuracy combining face recognition and other
biometric recognition .
 It will find efficiently without exhaustively searching the image.
 Face recognition systems are going to have widespread application in
smart environments.
23
06/04/2018
1. Steve Lawrence, C. Lee Giles , “Face Recognition: A Convolutional Neural Network Approach”, IEEE
transaction, St. Lucia, Australia.
2. David a brown, Ian craw, Julian lewthwaite, “Interactive Face retrieval using self organizing maps-A
SOM based approach to skin detection with application in real time systems”, IEEE 2008 conference,
Berlin, Germany.
3. Shahrin Azuan Nazeer, Nazaruddin Omar' and Marzuki Khalid, “Face Recognition System using
Artificial Neural Networks Approach”, IEEE - ICSCN 2007, MIT Campus, Anna University, Chennai,
India. Feb. 22-24, 2007. pp.420-425.
4. M. Prakash and M. Narasimha Murty, “Recognition Methods and Their NeuralNetwork Models”, IEEE
TRANSACTIONS ON NEURAL NETWORKS, VOL. 8, NO. 1, JANUARY 2005.
24
06/04/2018
25
06/04/2018

FACE RECOGNITION USING NEURAL NETWORK

  • 1.
    A SEMINAR ON FACE RECOGNITIONUSING NEURAL NETWORK Presented by- Soumyajit Sarkar(Roll No-16900315100) Tithi Dan(Roll No-16900315119) Mentor :- Prof. SubhamPramanik Department of Electronics & Communication Engineering
  • 2.
     Introduction  History Face recognition  Neural network  TECHNOLOGICAL IDEAS  ADVANTAGES  DISADVANTAGES  APPLICATIONS  CONCLUSIONS  REFERENCES 2 06/04/2018
  • 3.
     In thepresent scenario , there is great need to maintain information security or protection for physical property.  When credit and ATM cards are lost or stolen, an unauthorized user can often come up with the correct personal codes. 3 06/04/2018
  • 4.
     Face Recognitionis the fastest verification technology as it works with the most obvious individual i.e. The human face .  Information and property can be secured through verification of “true” individual identity.  It consists of unique shape analysis, pattern and positioning of facial features. 4 06/04/2018
  • 5.
     In 1960s,the first semi-automated system for facial recognition to locate the features(such as eyes, ears, nose and mouth) on the photographs.  In 1970s, Goldstein and Harmon used 21 specific subjective markers such as hair colour and lip thickness to automate the recognition.  In 1988, Kirby and Sirovich used standard linear algebra technique, to the face recognition. 5 06/04/2018
  • 6.
     Face recognitioninvolves comparing an image with a database of stored faces in order to identify the individual in that input image. 6  Face recognition involves comparing an image with a database of stored faces in order to identify the individual in that input image. 06/04/2018
  • 7.
     Face recognitiontechnology is the least intrusive. It works with the most obvious individual identifier-the human face.  It requires no physical interaction on behalf of the user.  It can use your existing hardware infrastructure , existing cameras and image capture devices will work with no problems. 7 06/04/2018
  • 8.
     It isa system of programs and data structures that approximates the operation of the human brain. 8 06/04/2018
  • 9.
     Adaptive learning:An ability to learn how to do task.  Self-Organization: Neural Network can create its own organisation.  Remarkable ability to derive meaning from complicated or imprecise data. 9 06/04/2018
  • 10.
     Here recognitionis performed by both Principal Component Analysis (PCA) and Back propagation Neural Networks(BPNN). All these processes are implemented for Face Recognition, based on the basic block diagram as shown in fig 1. BASIC BLOCK DIAGRAM(fig-1) Pre-processed input image Back Propagation Neural Network (BPNN) Principal Component Analysis (PCA) Classified Output Image 10 06/04/2018
  • 11.
     It normalizeand enhance the face image to improve the recognition performance. 11 06/04/2018
  • 12.
     PCA isa common statistical technique for finding the patterns in high dimensional data’s Feature extraction, also called Dimensionality Reduction. 12 06/04/2018
  • 13.
     Step 1:Partition face images into sub-patterns . 13 06/04/2018
  • 14.
     Step 2:Compute the expected contribution of each sub-pattern  Generate the Mean and Median faces for each person, and use these “virtual faces” as the probe set in training  Use the raw face-image sub-patterns as the gallery set in for training, and compute the PCA’s projection matrix on these gallery set  For each sample in the probe set, compute its similarity to the samples in corresponding gallery set 14 06/04/2018
  • 15.
     If asample from a sub-pattern’s probe set is correctly classified, the contribution of this sub-pattern is added by 1. 15 06/04/2018
  • 16.
    When an unknownface image comes in -  partition it into sub-patterns.  classify the unknown sample’s identity in each sub-pattern .  Incorporate the expected contribution and the classification result of all sub-patterns to generate the final classification result. 16 06/04/2018
  • 17.
     It trainsthe network to achieve a balance between the ability to respond correctly to the input patterns that are used for training & the ability to provide good response to the input that are similar.  It requires a dataset of the desired output for many input, making up the training set. 17 06/04/2018
  • 18.
     These arenecessarily Multilayer Perceptrons (MLPs).  MLPs: 1. Set of input layers 2. One or more hidden layers 3. Set of output layers 18 06/04/2018
  • 19.
     In anutshell, face recognition is done in this way- 19 06/04/2018
  • 20.
     Fastest securitymechanism.  No physical interaction .  more user friendly.  no extra learning process.  Simple, Fast & Easy to use.  Social acceptability. 20 06/04/2018
  • 21.
     Identical twinsattack  Requires straight on, natural expression  Affected by environment 21 06/04/2018
  • 22.
     criminals identificationin public location such as airport, Banks.  Building security  Credit card verification  Mobile phone unlocking 22 06/04/2018
  • 23.
     We haveto improve accuracy combining face recognition and other biometric recognition .  It will find efficiently without exhaustively searching the image.  Face recognition systems are going to have widespread application in smart environments. 23 06/04/2018
  • 24.
    1. Steve Lawrence,C. Lee Giles , “Face Recognition: A Convolutional Neural Network Approach”, IEEE transaction, St. Lucia, Australia. 2. David a brown, Ian craw, Julian lewthwaite, “Interactive Face retrieval using self organizing maps-A SOM based approach to skin detection with application in real time systems”, IEEE 2008 conference, Berlin, Germany. 3. Shahrin Azuan Nazeer, Nazaruddin Omar' and Marzuki Khalid, “Face Recognition System using Artificial Neural Networks Approach”, IEEE - ICSCN 2007, MIT Campus, Anna University, Chennai, India. Feb. 22-24, 2007. pp.420-425. 4. M. Prakash and M. Narasimha Murty, “Recognition Methods and Their NeuralNetwork Models”, IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 8, NO. 1, JANUARY 2005. 24 06/04/2018
  • 25.