Face Recognition Using Artificial Neural Network_final

25
Face Recognition using  Artificial Neural Network 

Transcript of Face Recognition Using Artificial Neural Network_final

Page 1: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 1/25

Face Recognition using

 Artificial Neural Network 

Page 2: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 2/25

Contents

►Problem specification

►Motivation

►Design

►Work done

►Results

►Future work 

►Demonstration

►References

Page 3: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 3/25

Problem Specification

To develop a face recognition system that:

►Takes a face image of a person as an input.

►Compares the face image of a person with theexisting face images that are already stored in thedatabase.

►Reports whether it is identified or not.

Page 4: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 4/25

Motivation

►Identity fraud is becoming a major concernfor all the governments around the globe

►Reliable methods of biometric personalidentification exists ,but these methods relyon the cooperation of the participants

►neural networks are good tool forclassification purposes

Page 5: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 5/25

Design

Image

Sampling 

Karhunen 

Loeve (KL)Transform

MultilayerPerceptron

ClassificationImage

Page 6: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 6/25

Image sampling

Page 7: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 7/25

KL Transform

►To reduce the dimensions of the imagevector

►Based on eigen values and correspondingeigen vectors.

Page 8: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 8/25

Multilayer perceptron

Page 9: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 9/25

Training a neural network 

►We train our neural network with a largesample of images.

►We wish to find the collection of weightsthat minimizes || TNET - T ACTUAL || .

Page 10: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 10/25

Testing

► After training is complete then the systemas a whole is ready to be used forrecognizing any given image.

►Testing image is used as an input to oursystem, the output of the system iscompared against the values stored in thedatabase.

►System reports whether a match ormismatch.

Page 11: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 11/25

Work Done

►Main concern in the project: Face recognition andnot face detection.

►Database of preprocessed images taken

CMU AMP Face Expression Database►contains 975 images of 13 subjects (75 images of each person)

► ‘bmp’ format with slightly varying poses, expressions etc

►converted into ‘pgm’ format using GIMP

► Separate java classes for K L transform

Multilayer Perceptron (MLP)

Training the MLP

Page 12: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 12/25

►Package named JAMA (Java matrix) used►Contains matrix operations like covariance, inverse,

transpose etc.

►Coding done in java. Reasons being:►To make application platform independent

►Java’s ability to handle large numbers 

Object oriented: to model real life situations

Page 13: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 13/25

►Neural net features: Number of input layer neurons: Number of Eigenvalues

Number of hidden layers: 1

Number of hidden layer neurons: 24(can be changed)

Number of output layer neurons: total number of subjects

Output given by neurons: 0 or 1

► Working

Training done with training images

 Validation done for the test images

 Appropriate message generated if subject is identified or notidentified

Page 14: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 14/25

RESULTS

►Different permutations tried for :

Hidden layer neurons

Output neurons

Form of outputs

Training cycles

Learning rate

►Done to bring error in an acceptable range

Page 15: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 15/25

► Satisfactory results obtained for following combination : Input neurons : selected Eigens

Hidden neurons : 24(can be changed)

Output neurons: total number of different subjects

Training cycles: 100000 Learning Rate: 0.3

Error obtained: 2.42E-4

► The system identified the subjects presented duringtraining

► For subjects not given during training : System refused toidentify

Page 16: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 16/25

FUTURE WORK 

►Face detection can be implemented

►Processing of image can be incorporated

Output of unidentified persons can bestored for future reference

► Ensemble of MLPs can be implemented

►Incremental learning can be implemented

Page 17: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 17/25

The mean Image

Page 18: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 18/25

DEMO

Page 19: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 19/25

 After training

Page 20: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 20/25

Selecting Image

Page 21: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 21/25

Match Found

Page 22: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 22/25

No Match

Page 23: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 23/25

References

[1] Steve Lawrence, C. Lee Giles, Ah Chung Tsoi, Andrew D. Back, Face 

Recognition: A Hybrid Neural Network Approach , Technical Report,UMIACS-TR-96-16 and CS-TR-3608, Institute for Advanced Computer

Studies, University of Maryland, 1996.

[2] Wendy S. Yambor, Analysis of PCA-based and Fisher discriminant- 

based image recognition algorithms , Technical Report CS-00-103,

Computer Science Department, Colorado State University, July 2000.

[3]Stuart Russel, Peter Norvig, Artificial Intelligence: A Modern 

 Approach , Pearson Education, 2nd Edition. 

Page 24: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 24/25

[4] Matthew A. Turk, Alex P. Pentland, Face Recognition Using Eigenfaces, Vision and Modeling Group , The Media Laboratory,Massachusetts Institute of Technology, 1991.

[5] W. Zhao, R. Chellappa, A. Rosenfeld, P.J. Phillips, Face Recognition: 

 A Literature Survey , ACM Computing Surveys, 2003, pp. 399-458.[6] T. De Bie, N. Cristianini, R. Rosipal, Eigenproblems in Pattern 

Recognition , Handbook of Computational Geometry for PatternRecognition, Computer Vision, Neurocomputing and Robotics, E. Bayro-Corrochano (editor), Springer-Verlag, Heidelberg, April 2004.

[7] Bai-Bo Zhang, Chang-Shui Zhang, Lower Bounds Estimation to KL Transform in Face Representation and Recognition , Proceedings of theFirst International Conference on Machine Learning and Cybernetics,Beijing, 4-5 November 2002.

Page 25: Face Recognition Using Artificial Neural Network_final

8/3/2019 Face Recognition Using Artificial Neural Network_final

http://slidepdf.com/reader/full/face-recognition-using-artificial-neural-networkfinal 25/25

[8]  An Introduction to Linear Algebra , :

http://www.cs.princeton.edu/introcs/95linear/  

[9] John Heaton , An Introduction to Neural Networks in Java  ,http://www.samspublishing.com 

[10] H.M. Deitel, P.J. Deitel, Java How to Program  , Pearson

Education,5th Edition