GENDER CLASSIFICATION USING SUPPORT VECTOR MACHINES

Size: px
Start display at page:

Download "GENDER CLASSIFICATION USING SUPPORT VECTOR MACHINES"

Transcription

1 GENDER CLASSIFICATION USING SUPPORT VECTOR MACHINES Ashwin Swaminathan ENEE633: Statistical and Neural Pattern Recognition Instructor : Prof. Rama Chellappa Project 2, Part (a) 1. INTRODUCTION Gender classification has been known as one of the most important tasks for humans [13]. Recently, the field of gender classification has caught considerable attention due to its emerging applications in the area of human computer interactions systems [14] and other applications. The problem of gender classification has been widely studied in psychology where they try to model the human recognition system to classify gender. Over the past few years, there have been a few learning based algorithms proposed for gender classification. Most of these methods use neural networks for gender classification [16, 17]. In these works, the authors perform tests over 30 images and show that the neural network can help determine gender from 8x8 images with an average accuracy of 92%. RBF type networks have been widely used for this purpose. The use of Support vector machines for gender classification was proposed by Moghaddam et al. in [14]. It is well known that SVM can provide superior performace and hence it has been widely used in many classification problems. Specifically, it has been very helpful in cases where it is difficult to estimate the density model such as in classification of images e.g. face recognition, etc. The support vector machines employing kernel functions have been useful in cases where the data are not linearly separable. These kernel functions help in transforming the data to a higher dimensional space where they can be distinguished easily. In this paper, the authors use SVM for classification. They use hair-less images from around 1800 faces each in a 21 x 12 format and show that they can achieve an average

2 performance accuracy of around 96% using a support vector machine employing radial basis functions as its kernel. In this project, we explore further on the use of support vector machines for the problem of gender classification. In particular, we study the performance of the SVM under different types of kernel functions such as linear, polynomial, radial basis functions and sigmoid. We then consider the effect of feature generation techniques such as principal component analysis and fisher linear discriminant and compare the results with those obtained using the image pixel intensity values as features. 2. SUPPORT VECTOR MACHINES FOR CLASSIFICATION The SVM is a learning algorithm for classification. It tries to find the optimal separating hyperplane such that the expected classification error for unseen patterns is minimized [18]. Thus, the SVM has very good generalization performance. More precisely, given a set of training samples x i and the corresponding decision values y i { 1, 1}, the SVM aims at finding the best separating hyperplane given by the equation w T x + b that maximizes the distance between the two classes. The set of points for which w T x + b 1 are set to belong to one class and those points for which w T x + b 1 are classified into another. The goal of the optimizing function is to maximize the distance between the two hyper-planes given by 1. This problem can be equivalently formulated of the form w T w minimize w T w subject to the constraint that y i (x T i w + b) 1 i The solution to this optimization problem can be easily obtained when the data are linearly separable. However, complications arise when the data are non-separable. In this case, the SVM handles non-separable data by introducing slack variables ξ i. In this case, the optimization problem can be reformulated of the form minimize w C( ξ i ) k

3 subject to the constraint that x T i w + b 1 ξ i y i = 1 x T i w + b 1 + ξ i y i = 1 ξ i 0 (1) The primal problem as stated above can be converted to its equivalent dual form maximize α i 1 α i α j y i y j x T i x j 2 subject to 0 α i C and i α iy i = 0. In general, the term x T i x j can be rewritten in terms of the kernel functions as K(x i, x j ). In this case, the problem can be equivalently understood in terms of projecting the input data into a higher dimensional space where they are separated using to parallel hyperplanes. There exist many popular kernel functions that have been widely used for classification. The polynomial kernel function is of the form K(x i, x j ) = (1 + x T i x j ) p i j and the radial basis function is given by K(x i, x j ) = exp( γ x j x i 2 ) (2) By using different kernel functions, SVM can implement a wide variety of learning algorithms. It is well known that the SVM has a great potential to perform well. However, the performance of the SVM is very closely tied to the choice of the optimal kernel functions. There has been a lot of research over the last few years on algorithms to help choose the exact type of kernel for a given problem with a certain set of features. Most of these methods are based on simple heuristics based on the knowledge of the input data and there has not been any standardized method to obtain the best kernel. Hence, the choice of the optimal kernel has reduced to a trial and error procedure in most scenarios. 3. SIMULATION RESULTS In this section, we present the simulation results for the gender classification problem with support vector machines.

4 3.1. Simulation setup We use the gender classification database given in class that has been adopted from [15]. This database has a total of 265 images. The first 150 of them are male and the remaining are female subjects. It has two images per person each of dimension These image have been cropped and aligned to produce only the face region. Some sample images from the database are shown in Fig. 1. The image intensity values were used as features for classification. In the following sections, we present results with the SVM tested under different kinds of scenario and under different kinds of kernel types. We start with the simple linear kernel based SVM. Fig. 1. Sample inputs for the Gender classification Problem 3.2. Simulation Results with Linear Support Vector Machine In our study, we first use the linear SVM kernel. We use 30 images from each gender for training and test it with 30 different images from each class. In each case, the number of female faces classified as male and the number of male classified as female are noted. The experiment was repeated numerous times by choosing a different set of images in training and testing set and the average performance was computed. The average confusion matrix is shown in Table 1. We notice that around 2.4% of the males are wrongly classified as females. In the case of female images, we observe a larger error and around 29.4% of the female faces are classified as males. On closer examination, we observe that there are 46 different support faces 24 of which are males

5 Table 1. Average Confusion Matrix when classified with the SVM with Linear kernel Male Female Male 97.60% 2.40% Female 29.60% 70.40% and the remaining 22 are females. Some of the support faces are shown in Fig. 2. Fig. 2. Sample Support faces obtained using Linear SVM Next, we study the performance as the number of training images in each class is increased. In all cases, we fix the number of testing images to 30. The percentage accuracies for male and female face classification are shown in Fig. 3. These results again indicate that there is considerable more error in classifying female faces. Further, we notice that the percentage accuracy does not improve much on increasing the size of training set. In Table 2, we show the average number of support faces when the number of training images are increased. We notice that when the number of training images is small, almost all the faces become support faces. Thus, in this case all the images in the training set has a very high influence on the performance during testing.

6 100 Classification Accuracy with Linear SVM Percentage Accuracy Male Female Average No. of Training images Fig. 3. Probability of Correct classification for SVM with Linear kernel with the number of training images (a) Males shown in blue (b) Females in red (c) Average performance shown in black. Table 2. Average Number of Support Faces as a function of Number of training images Num Train (per class) No. Male SVs No. Female SVs Simulation Results with SVM with Polynomial kernel We noticed from the previous subsection that the SVM with the linear kernel performed quite well for classification of males. But, it did not do quite well with females and gave a lot of errors. In most cases, the percentage accuracy for females was only around 75% and the performance did not improve much on increasing the size of the training set. Next, we study the performance with the polynomial kernel is used. The simple polynomial kernel can be expressed as K(x i, x j ) = (1+x T i x j ) p. First, we study the performance as the degree of the polynomial kernel is changed. The results are shown in Fig. 4. We observe that the classification performance improves as the degree of the polynomial kernel is increased initially and the polynomial kernel with

7 degree 2 performs better than the kernel with degree 1. Further, we notice that the performance almost saturates and does not change much as the degree is further increased. In this case, our results indicate that we have on an average about 16.2 male support faces and around 14.3 female support faces. Hence, the total number of faces in the boundary is 30.5 which is much less than 44 which was observed for the linear kernel. This indicates that the polynomial kernel is able to better fit the data Male Female Average Percentage Accuracy Degree of Polynomial kernel Fig. 4. Probability of Correct classification for various degrees of polynomial kernels for the case (a) Males shown in blue (b) Females in red (c) Average performance shown in black. Next, we study the performance of the polynomial kernel (with degree 3) as the number of training images is increased. The results are shown in Fig. 5. We notice that the performance improves as the number of training images is increased. We also observe that the performance in this case is much better compared to the linear SVM case explored in the previous subsection. The mediocre performance obtained with the linear SVM seems to indicate that a linear decision surface is not able to capture the properties of the dataset Simulation Results with SVM with RBF kernel In this subsection, we study the performance with the radial basis function kernel. The RBF kernel can be expressed as K(x i, x j ) = (exp( γ x i x j 2 )). First, we study the performance as the value of γ is changed. The results are shown in Fig. 6. We

8 96 Percentage Accuracies for SVM with Polynomial kernel with degree Percentage Accuracy Male Female Average Number of Training images Fig. 5. Probability of Correct classification for SVM with polynomial kernel with the number of training images for the case of (a) Males shown in blue (b) Females in red (c) Average performance shown in black. observe that the classification performance does not change much as the value of γ is changed. In this case, we use 30 images under each gender for training and 30 for testing. We notice that the average performance of the SVM with the RBF kernel is not as good as the SVM with polynomial kernel. This seems to indicate that the polynomial kernel is able to better fit the data in this case. Further, our results indicate that we have on an average about 19.8 male support faces and around 18.6 female support faces. Hence, the total number of faces in the boundary is higher than the polynomial kernel case where it was about This indicates that the polynomial kernel is able to better fit the data. Next, we study the performance of the RBF kernel (with γ = 0.5) as the number of training images is increased. The results are shown in Fig. 7. We notice that the performance does not change much as the size of the training set is increased Simulation Results with SVM with Sigmoid kernel Finally, we study the performance with the sigmoid kernel. The sigmoid kernel can be expressed as K(x i, x j ) = tanh(γ x T i x j + c). First, we study the performance as the value of γ is changed. The results are shown in Fig. 8. We observe that the classification performance changes drastically as the value of γ is changed. As the value

9 100 Percentage Accuracy for SVM with RBF kernel Percentage Accuracy Male Female Average γ Fig. 6. Probability of Correct classification for SVM with RBF kernel with the parameter γ for the case of (a) Males shown in blue (b) Females in red (c) Average performance shown in black. is increased, we notice that the classification performance for males decreases and the corresponding performance for females increases. The average performance is almost a constant. Thus, on increasing the value of γ the sigmoid kernel is able to better learn the female faces. We also see that the performance of the sigmoid kernel is not as good as the polynomial kernel which seems to provide the best performance. Further, our results indicate that we have on an average about 29.8 male support faces and around 28.6 female support faces. Hence, the total number of faces in the boundary is higher than the polynomial kernel case where it was about This indicates that the polynomial kernel is able to better fit the data. Next, we study the performance of the RBF kernel (with γ = 1.6) as the number of training images is increased. The results are shown in Fig. 9. We notice that the performance does not change much as the size of the training set is increased Comparison under Different Kernel types In this subsection, we compare the performance for all the four types of kernels considered, namely, linear, polynomial, RBF and sigmoid. The comparison results for 30 training images are shown in Table 3. The corresponding average number of support vectors are shown in Table 4. We notice from the

10 100 Results for RBF kernel Percentage Accuracy Male Female Average No. of Training images per class Fig. 7. Probability of Correct classification for SVM with RBF kernel with the number of training images for the case of (a) Males shown in blue (b) Females in red (c) Average performance shown in black. results that the SVM with the polynomial kernel provides the best classification accuracies. Thus, it is able to better represent the data. This is also evident from Table 4 where it is clear that the SVM with polynomial kernel uses the least number of support vectors. Thus, after elaborate and detailed comparison studies, we notice that the polynomial kernel performs the best for gender classification. Table 3. Average Accuracies for SVM classifier with different kernels Classifier % Overall Male Female Linear Polynomial with degree Radial Basis Function Sigmoid GENDER CLASSIFICATION USING PRINCIPAL COMPONENT ANALYSIS The Principal Component Analysis method functions by projecting a face onto a multi-dimensional feature space that spans the gamut of human faces. A set of basis images is extracted from the database presented to the system by Eigenvalue-Eigenvector decomposition. Any face in the feature space is then

11 Percentage Accuracies with Sigmoid Kernel Male Female Average Percentage Accuracy γ Fig. 8. Probability of Correct classification for SVM with sigmoid kernel with the parameter γ for the case of (a) Males shown in blue (b) Females in red (c) Average performance shown in black. Table 4. Average Number of Support Vectors for SVM classifier with different kernels Classifier % Male Female Linear Polynomial with degree Radial Basis Function Sigmoid characterized by a weight vector obtained by projecting it onto the set of basis images. When a new face is presented to the system, its weight vector is calculated and a SVM based classifier is used for classification. The full details of the algorithm can be found in [1]. In this section, we present the results for the gender classification algorithm after PCA was done on the data set. The correlation matrix of the input data was first computed and its eigenvectors were found. The training data was then projected on to these basis vectors and the projection coefficients were estimated. These coefficients were then used for classification. The SVM was used for classifying the data and all the four kernel functions were tried. The results are shown in Table 5. Here, we show the results both with and without PCA for comparison. We notice that the classification performance improves after doing PCA. The improvement is especially pronounced for female data where it is around 8-9% in most cases. This indicates that the female faces are better

12 Male Female Average Percentage Accuracies for Sigmoid Kernel Percentage Accuracy No. of training images for each class Fig. 9. Probability of Correct classification for SVM with sigmoid kernel with the number of training images for the case of (a) Males shown in blue (b) Females in red (c) Average performance shown in black. represented and classified accurately when the projection coefficients are used in classification. There is also an overall enhancement of about 1-2%. Again, we notice that the polynomial kernel with degree 3 performs best for classification. Our results indicate that the polynomial kernel is better able to fit the face data. Table 5. Average Accuracies for SVM classifier with different kernels with and without doing PCA on the test data Features SVM Kernel Type % Overall Male Female Image Pixels Linear SVM Polynomial with degree Radial Basis Function Sigmoid PCA coefficients Linear SVM Polynomial with degree Radial Basis Function Sigmoid

13 5. GENDER CLASSIFICATION USING FISHER LINEAR DISCRIMINANT The PCA takes advantage of the fact that, under admittedly idealized conditions, the variation within class lies in a linear subspace of the image space [2]. Hence, the classes are convex, and, therefore, linearly separable. One can perform dimensionality reduction using linear projection and still preserve linear separability. However, when the variation in classes is large, there is a lot of variance in the observed features and this leads to wrong classification results. This problem can be offset by considering a modified set of optimization leading to a different set of basis vectors. Instead of finding the best set of basis vectors to minimize the representation error, a modified cost function is considered to improve classification accuracy. More specifically, the basic idea behind Fisher Linear discriminant analysis is to find the best set of vectors that can minimize the intra cluster variability while maximizing the inter cluster distances. The simulation results for FLD are shown in Table 6. Here, we show the results using PCA and direct image features also for comparison purposes. We use 60 images for training and test it with 60 other images to generate our simulation results. Out of these, 30 are male and the remaining 30 are females. These 60 images are chosen in a random order and a different set of faces is chosen each time. The experiment is repeated over a 100 times and the average performance is recorded as shown in the table. In the case of FLD, we first do a PCA to reduce the dimensionality of the input space to 60. The within class scatter and the between class scatter matrix is computed. The top 5 eigenvectors corresponding to the solution of the generalized eigenvalue problem are found. The input data is then projected on to these top 5 fisher basis. These 5 coefficients are used for classification. The results of our study and the overall performance is shown in Table 6. We notice a slight improvement in performance with compared with directly using the image pixel intensities as features. However, it to be noted that while using the image pixel values, the length of our feature vector was around 9750 compared to 5 in the case of FLD. Thus, the FLD was able to achieve a much better performance even under a highly reduced dimensional space. Again, in this case like in PCA, we notice that the average classification accuracy for females improved tremendously in comparison with using image pixel values as features. This is expected in the light of the fact that we are using better set of features by projecting the input faces on to a set of basis vectors. Our results also indicate that the SVM with polynomial kernel with degree 3 seems to give the best results.

14 Table 6. Average Accuracies for SVM classifier with different kernels and different features Features SVM Kernel Type % Overall Male Female Image Pixels Linear SVM Polynomial with degree Radial Basis Function Sigmoid PCA coefficients Linear SVM Polynomial with degree Radial Basis Function Sigmoid FLD coefficients Linear SVM Polynomial with degree Radial Basis Function Sigmoid CONCLUSIONS In this project, we consider the problem of gender classification and use the Support vector machines for our study. As a first step for gender classification, we use the image pixel intensity values as features. These are then fed into the SVM with different types of kernel functions and the classification error and accuracy are noted in each case. Our results indicate that even on using the image pixel values as features, the SVM has the potential to segregate the images into the two classes of male and female with a high accuracy. In particular, we explored the use of four different kernels namely linear, polynomial, radial basis function and sigmoid. The performance under different types of parameters was studied and tabulated. Our results indicate that that the polynomial kernel with degree 3 gives the best performance compared to all the other kernels (with different parameter settings). Next, we consider feature dimension reduction methods. More specifically, we test the performance of the SVM classifier when the principal component coefficients and the fisher linear discriminants are used as features. Again, in this case, our elaborated and detailed comparisons indicate that the FLD with polynomial kernel seems to give the best results.

15 7. REFERENCES [1] M. Turk, A. Pentland, Eigenfaces for recognition, Journal of Cognitive Neuroscience, vol. 3, pp 72-86, [2] P. Belhumeur, J. Hespanha, and D. Kriegman, Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection, IEEE Trans. PAMI, vol. 19, pp , [3] K. Etemad and R. Chellappa, Discriminant Analysis for Recognition of Human Face Images, Journal of Optical Society of America A, pp , [4] W. Zhao, R. Chellappa, A. Rosenfeld, and J. Phillips, Face Recognition: A Literature Survey, to appear ACM computing surveys, 2003 [5] S.K.Zhou and R.Chellappa, Multiple-Exemplar Discriminant Analysis for Face Recognition, ICIP. [6] M.S.Bartlett, J.R.Movellan, T.R.Sejnowski, Face recognition by Independent Component Analysis, IEEE Trans. on Neural Networks, Vol. 13, No. 6, Nov [7] B.Moghaddam, T.Jebara, A. Pentland, Bayesian Face recogition, MERL research report, Feb [8] B.Moghaddam, T.Jebara, A. Pentland, Bayesian Modeling of Facial Similarity, Adv. in Neural Info. Processing Systems 11, MIT Press, [9] B.Moghaddam, A.Pentland, Probabilisitc Visual Learning for Object Representation, Early Visual Learning, Oxford University Press, [10] S.Z.Li, J. Lu, Face Recognition Using Nearest Feature Line Method, IEEE Trans. on Neural Networks, Vol. 10, No. 2, [11] M.Ramachandran, S.K.Zhou, R.Chellappa, D.Jhalani, Methods to convert smiling face to neutral face with applications to face reognition, IEEE ICASSSP, March [12] R. Chellappa, C. Wilson, and S. Sirohey, Human and Machine Recognition of Faces: A Survey, Proceedings of IEEE, vol. 83, pp , [13] A. B. A. Graf, and F. A. Wichmann, Gender Classification of Human Faces, Technical report on MERL.

16 [14] B. Moghaddam, and M.-H. Yang, Learning Gender with Support Faces, in IEEE Trans. on Pattern Analysis and Machine Intelligence, Vol. 24, NO. 7, July [15] A.M. Martinez and R. Benavente. The AR Face Database. CVC Technical Report no. 24, June [16] B. A. Gollomb, D. T. Lawrence, and T. J. Sejnowski, Sexnet: A Neural Network Identifies Sex from Human Faces, Advances in Neural Information Processing Systems, pp , [17] G. W. Cottrell, Empath: Face, Emotion and Gender Recognition Using Holons, Advances in Neural Information Processing Systems, pp , [18] C. J. C. Burges, A Tutorial on Support Vector Machines for Pattern Recognition, Kluwer Academic Publishers, Boston.

FACE RECOGNITION USING SUPPORT VECTOR MACHINES

FACE RECOGNITION USING SUPPORT VECTOR MACHINES FACE RECOGNITION USING SUPPORT VECTOR MACHINES Ashwin Swaminathan ashwins@umd.edu ENEE633: Statistical and Neural Pattern Recognition Instructor : Prof. Rama Chellappa Project 2, Part (b) 1. INTRODUCTION

More information

Linear Discriminant Analysis in Ottoman Alphabet Character Recognition

Linear Discriminant Analysis in Ottoman Alphabet Character Recognition Linear Discriminant Analysis in Ottoman Alphabet Character Recognition ZEYNEB KURT, H. IREM TURKMEN, M. ELIF KARSLIGIL Department of Computer Engineering, Yildiz Technical University, 34349 Besiktas /

More information

Image Processing and Image Representations for Face Recognition

Image Processing and Image Representations for Face Recognition Image Processing and Image Representations for Face Recognition 1 Introduction Face recognition is an active area of research in image processing and pattern recognition. Since the general topic of face

More information

A Hierarchical Face Identification System Based on Facial Components

A Hierarchical Face Identification System Based on Facial Components A Hierarchical Face Identification System Based on Facial Components Mehrtash T. Harandi, Majid Nili Ahmadabadi, and Babak N. Araabi Control and Intelligent Processing Center of Excellence Department of

More information

Image-Based Face Recognition using Global Features

Image-Based Face Recognition using Global Features Image-Based Face Recognition using Global Features Xiaoyin xu Research Centre for Integrated Microsystems Electrical and Computer Engineering University of Windsor Supervisors: Dr. Ahmadi May 13, 2005

More information

FACE RECOGNITION BASED ON GENDER USING A MODIFIED METHOD OF 2D-LINEAR DISCRIMINANT ANALYSIS

FACE RECOGNITION BASED ON GENDER USING A MODIFIED METHOD OF 2D-LINEAR DISCRIMINANT ANALYSIS FACE RECOGNITION BASED ON GENDER USING A MODIFIED METHOD OF 2D-LINEAR DISCRIMINANT ANALYSIS 1 Fitri Damayanti, 2 Wahyudi Setiawan, 3 Sri Herawati, 4 Aeri Rachmad 1,2,3,4 Faculty of Engineering, University

More information

Disguised Face Identification Based Gabor Feature and SVM Classifier

Disguised Face Identification Based Gabor Feature and SVM Classifier Disguised Face Identification Based Gabor Feature and SVM Classifier KYEKYUNG KIM, SANGSEUNG KANG, YUN KOO CHUNG and SOOYOUNG CHI Department of Intelligent Cognitive Technology Electronics and Telecommunications

More information

Learning to Recognize Faces in Realistic Conditions

Learning to Recognize Faces in Realistic Conditions 000 001 002 003 004 005 006 007 008 009 010 011 012 013 014 015 016 017 018 019 020 021 022 023 024 025 026 027 028 029 030 031 032 033 034 035 036 037 038 039 040 041 042 043 044 045 046 047 048 049 050

More information

Multidirectional 2DPCA Based Face Recognition System

Multidirectional 2DPCA Based Face Recognition System Multidirectional 2DPCA Based Face Recognition System Shilpi Soni 1, Raj Kumar Sahu 2 1 M.E. Scholar, Department of E&Tc Engg, CSIT, Durg 2 Associate Professor, Department of E&Tc Engg, CSIT, Durg Email:

More information

Semi-Supervised PCA-based Face Recognition Using Self-Training

Semi-Supervised PCA-based Face Recognition Using Self-Training Semi-Supervised PCA-based Face Recognition Using Self-Training Fabio Roli and Gian Luca Marcialis Dept. of Electrical and Electronic Engineering, University of Cagliari Piazza d Armi, 09123 Cagliari, Italy

More information

On Modeling Variations for Face Authentication

On Modeling Variations for Face Authentication On Modeling Variations for Face Authentication Xiaoming Liu Tsuhan Chen B.V.K. Vijaya Kumar Department of Electrical and Computer Engineering, Carnegie Mellon University, Pittsburgh, PA 15213 xiaoming@andrew.cmu.edu

More information

An Integrated Face Recognition Algorithm Based on Wavelet Subspace

An Integrated Face Recognition Algorithm Based on Wavelet Subspace , pp.20-25 http://dx.doi.org/0.4257/astl.204.48.20 An Integrated Face Recognition Algorithm Based on Wavelet Subspace Wenhui Li, Ning Ma, Zhiyan Wang College of computer science and technology, Jilin University,

More information

IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 14, NO. 1, JANUARY Face Recognition Using LDA-Based Algorithms

IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 14, NO. 1, JANUARY Face Recognition Using LDA-Based Algorithms IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 14, NO. 1, JANUARY 2003 195 Brief Papers Face Recognition Using LDA-Based Algorithms Juwei Lu, Kostantinos N. Plataniotis, and Anastasios N. Venetsanopoulos Abstract

More information

Chakra Chennubhotla and David Koes

Chakra Chennubhotla and David Koes MSCBIO/CMPBIO 2065: Support Vector Machines Chakra Chennubhotla and David Koes Nov 15, 2017 Sources mmds.org chapter 12 Bishop s book Ch. 7 Notes from Toronto, Mark Schmidt (UBC) 2 SVM SVMs and Logistic

More information

Robot Learning. There are generally three types of robot learning: Learning from data. Learning by demonstration. Reinforcement learning

Robot Learning. There are generally three types of robot learning: Learning from data. Learning by demonstration. Reinforcement learning Robot Learning 1 General Pipeline 1. Data acquisition (e.g., from 3D sensors) 2. Feature extraction and representation construction 3. Robot learning: e.g., classification (recognition) or clustering (knowledge

More information

SINGLE-SAMPLE-PER-PERSON-BASED FACE RECOGNITION USING FAST DISCRIMINATIVE MULTI-MANIFOLD ANALYSIS

SINGLE-SAMPLE-PER-PERSON-BASED FACE RECOGNITION USING FAST DISCRIMINATIVE MULTI-MANIFOLD ANALYSIS SINGLE-SAMPLE-PER-PERSON-BASED FACE RECOGNITION USING FAST DISCRIMINATIVE MULTI-MANIFOLD ANALYSIS Hsin-Hung Liu 1, Shih-Chung Hsu 1, and Chung-Lin Huang 1,2 1. Department of Electrical Engineering, National

More information

Stepwise Nearest Neighbor Discriminant Analysis

Stepwise Nearest Neighbor Discriminant Analysis Stepwise Nearest Neighbor Discriminant Analysis Xipeng Qiu and Lide Wu Media Computing & Web Intelligence Lab Department of Computer Science and Engineering Fudan University, Shanghai, China xpqiu,ldwu@fudan.edu.cn

More information

A KERNEL MACHINE BASED APPROACH FOR MULTI- VIEW FACE RECOGNITION

A KERNEL MACHINE BASED APPROACH FOR MULTI- VIEW FACE RECOGNITION A KERNEL MACHINE BASED APPROACH FOR MULI- VIEW FACE RECOGNIION Mohammad Alwawi 025384 erm project report submitted in partial fulfillment of the requirement for the course of estimation and detection Supervised

More information

Bahar HATİPOĞLU 1, and Cemal KÖSE 1

Bahar HATİPOĞLU 1, and Cemal KÖSE 1 Gender Recognition from Face Images Using PCA and LBP Bahar HATİPOĞLU 1, and Cemal KÖSE 1 1 Department of Computer Engineering, Karadeniz Technical University, Trabzon, Turkey baharhatipoglu@ktu.edu.tr,

More information

Linear Discriminant Analysis for 3D Face Recognition System

Linear Discriminant Analysis for 3D Face Recognition System Linear Discriminant Analysis for 3D Face Recognition System 3.1 Introduction Face recognition and verification have been at the top of the research agenda of the computer vision community in recent times.

More information

Performance Evaluation of PCA and LDA for Face Recognition

Performance Evaluation of PCA and LDA for Face Recognition Performance Evaluation of PCA and LDA for Face Recognition S. K. Hese, M. R. Banwaskar Department of Electronics & Telecommunication, MGM s College of Engineering Nanded Near Airport, Nanded, Maharashtra,

More information

Enhanced (PC) 2 A for Face Recognition with One Training Image per Person

Enhanced (PC) 2 A for Face Recognition with One Training Image per Person Enhanced (PC) A for Face Recognition with One Training Image per Person Songcan Chen, *, Daoqiang Zhang Zhi-Hua Zhou Department of Computer Science Engineering Nanjing University of Aeronautics Astronautics,

More information

Gender Classification of Face Images: The Role of Global and Feature-Based Information

Gender Classification of Face Images: The Role of Global and Feature-Based Information Gender Classification of Face Images: The Role of Global and Feature-Based Information Samarasena Buchala, Neil Davey, Ray J.Frank, Tim M.Gale,2, Martin J.Loomes, Wanida Kanargard Department of Computer

More information

Recognition: Face Recognition. Linda Shapiro EE/CSE 576

Recognition: Face Recognition. Linda Shapiro EE/CSE 576 Recognition: Face Recognition Linda Shapiro EE/CSE 576 1 Face recognition: once you ve detected and cropped a face, try to recognize it Detection Recognition Sally 2 Face recognition: overview Typical

More information

Face Recognition Based on LDA and Improved Pairwise-Constrained Multiple Metric Learning Method

Face Recognition Based on LDA and Improved Pairwise-Constrained Multiple Metric Learning Method Journal of Information Hiding and Multimedia Signal Processing c 2016 ISSN 2073-4212 Ubiquitous International Volume 7, Number 5, September 2016 Face Recognition ased on LDA and Improved Pairwise-Constrained

More information

A New Multi Fractal Dimension Method for Face Recognition with Fewer Features under Expression Variations

A New Multi Fractal Dimension Method for Face Recognition with Fewer Features under Expression Variations A New Multi Fractal Dimension Method for Face Recognition with Fewer Features under Expression Variations Maksud Ahamad Assistant Professor, Computer Science & Engineering Department, Ideal Institute of

More information

Face Identification by a Cascade of Rejection Classifiers

Face Identification by a Cascade of Rejection Classifiers Boston U. Computer Science Tech. Report No. BUCS-TR-2005-022, Jun. 2005. To appear in Proc. IEEE Workshop on Face Recognition Grand Challenge Experiments, Jun. 2005. Face Identification by a Cascade of

More information

CHAPTER 3 PRINCIPAL COMPONENT ANALYSIS AND FISHER LINEAR DISCRIMINANT ANALYSIS

CHAPTER 3 PRINCIPAL COMPONENT ANALYSIS AND FISHER LINEAR DISCRIMINANT ANALYSIS 38 CHAPTER 3 PRINCIPAL COMPONENT ANALYSIS AND FISHER LINEAR DISCRIMINANT ANALYSIS 3.1 PRINCIPAL COMPONENT ANALYSIS (PCA) 3.1.1 Introduction In the previous chapter, a brief literature review on conventional

More information

FACE IDENTIFICATION ALGORITHM BASED ON MESH-DERIVED SYNTHETIC LINEAR DESCRIPTORS

FACE IDENTIFICATION ALGORITHM BASED ON MESH-DERIVED SYNTHETIC LINEAR DESCRIPTORS computing@tanet.edu.te.ua www.tanet.edu.te.ua/computing ISSN 1727-6209 International Scientific Journal of Computing FACE IDENTIFICATION ALGORITHM BASED ON MESH-DERIVED SYNTHETIC LINEAR DESCRIPTORS R.

More information

NIST. Support Vector Machines. Applied to Face Recognition U56 QC 100 NO A OS S. P. Jonathon Phillips. Gaithersburg, MD 20899

NIST. Support Vector Machines. Applied to Face Recognition U56 QC 100 NO A OS S. P. Jonathon Phillips. Gaithersburg, MD 20899 ^ A 1 1 1 OS 5 1. 4 0 S Support Vector Machines Applied to Face Recognition P. Jonathon Phillips U.S. DEPARTMENT OF COMMERCE Technology Administration National Institute of Standards and Technology Information

More information

Robust Face Recognition via Sparse Representation Authors: John Wright, Allen Y. Yang, Arvind Ganesh, S. Shankar Sastry, and Yi Ma

Robust Face Recognition via Sparse Representation Authors: John Wright, Allen Y. Yang, Arvind Ganesh, S. Shankar Sastry, and Yi Ma Robust Face Recognition via Sparse Representation Authors: John Wright, Allen Y. Yang, Arvind Ganesh, S. Shankar Sastry, and Yi Ma Presented by Hu Han Jan. 30 2014 For CSE 902 by Prof. Anil K. Jain: Selected

More information

Eigenfaces and Fisherfaces A comparison of face detection techniques. Abstract. Pradyumna Desale SCPD, NVIDIA

Eigenfaces and Fisherfaces A comparison of face detection techniques. Abstract. Pradyumna Desale SCPD, NVIDIA Eigenfaces and Fisherfaces A comparison of face detection techniques Pradyumna Desale SCPD, NVIDIA pdesale@nvidia.com Angelica Perez Stanford University pereza77@stanford.edu Abstract In this project we

More information

Gender Classification of Human Faces

Gender Classification of Human Faces Presented at: BMCV*02, Tübingen, FRG, 22-24 November 2002. Published in: Biologically Motivated Computer Vision, eds. H.H. Bülthoff, S.-W. Lee, T.A. Poggio and C. Wallraven, LNCS 2525, pages 491-1, 2002,

More information

All lecture slides will be available at CSC2515_Winter15.html

All lecture slides will be available at  CSC2515_Winter15.html CSC2515 Fall 2015 Introduc3on to Machine Learning Lecture 9: Support Vector Machines All lecture slides will be available at http://www.cs.toronto.edu/~urtasun/courses/csc2515/ CSC2515_Winter15.html Many

More information

12 Classification using Support Vector Machines

12 Classification using Support Vector Machines 160 Bioinformatics I, WS 14/15, D. Huson, January 28, 2015 12 Classification using Support Vector Machines This lecture is based on the following sources, which are all recommended reading: F. Markowetz.

More information

Hybrid Face Recognition and Classification System for Real Time Environment

Hybrid Face Recognition and Classification System for Real Time Environment Hybrid Face Recognition and Classification System for Real Time Environment Dr.Matheel E. Abdulmunem Department of Computer Science University of Technology, Baghdad, Iraq. Fatima B. Ibrahim Department

More information

Two-View Face Recognition Using Bayesian Fusion

Two-View Face Recognition Using Bayesian Fusion Proceedings of the 2009 IEEE International Conference on Systems, Man, and Cybernetics San Antonio, TX, USA - October 2009 Two-View Face Recognition Using Bayesian Fusion Grace Shin-Yee Tsai Department

More information

Kernel Methods & Support Vector Machines

Kernel Methods & Support Vector Machines & Support Vector Machines & Support Vector Machines Arvind Visvanathan CSCE 970 Pattern Recognition 1 & Support Vector Machines Question? Draw a single line to separate two classes? 2 & Support Vector

More information

Dimension Reduction CS534

Dimension Reduction CS534 Dimension Reduction CS534 Why dimension reduction? High dimensionality large number of features E.g., documents represented by thousands of words, millions of bigrams Images represented by thousands of

More information

Face Recognition Based On Granular Computing Approach and Hybrid Spatial Features

Face Recognition Based On Granular Computing Approach and Hybrid Spatial Features Face Recognition Based On Granular Computing Approach and Hybrid Spatial Features S.Sankara vadivu 1, K. Aravind Kumar 2 Final Year Student of M.E, Department of Computer Science and Engineering, Manonmaniam

More information

Local Similarity based Linear Discriminant Analysis for Face Recognition with Single Sample per Person

Local Similarity based Linear Discriminant Analysis for Face Recognition with Single Sample per Person Local Similarity based Linear Discriminant Analysis for Face Recognition with Single Sample per Person Fan Liu 1, Ye Bi 1, Yan Cui 2, Zhenmin Tang 1 1 School of Computer Science and Engineering, Nanjing

More information

Training Algorithms for Robust Face Recognition using a Template-matching Approach

Training Algorithms for Robust Face Recognition using a Template-matching Approach Training Algorithms for Robust Face Recognition using a Template-matching Approach Xiaoyan Mu, Mehmet Artiklar, Metin Artiklar, and Mohamad H. Hassoun Department of Electrical and Computer Engineering

More information

Vol. 4, No. 1 Jan 2013 ISSN Journal of Emerging Trends in Computing and Information Sciences CIS Journal. All rights reserved.

Vol. 4, No. 1 Jan 2013 ISSN Journal of Emerging Trends in Computing and Information Sciences CIS Journal. All rights reserved. An Automatic Face Detection and Gender Classification from Color Images using Support Vector Machine 1 Md. Hafizur Rahman, 2 Suman Chowdhury, 3 Md. Abul Bashar 1, 2, 3 Department of Electrical & Electronic

More information

LECTURE 5: DUAL PROBLEMS AND KERNELS. * Most of the slides in this lecture are from

LECTURE 5: DUAL PROBLEMS AND KERNELS. * Most of the slides in this lecture are from LECTURE 5: DUAL PROBLEMS AND KERNELS * Most of the slides in this lecture are from http://www.robots.ox.ac.uk/~az/lectures/ml Optimization Loss function Loss functions SVM review PRIMAL-DUAL PROBLEM Max-min

More information

Feature Search. Multiscale Head Search

Feature Search. Multiscale Head Search Sex with Support Vector Machines Baback Moghaddam Mitsubishi Electric Research Laboratory Cambridge MA 02139, USA baback@merl.com Ming-Hsuan Yang University of Illinois at Urbana-Champaign Urbana, IL 61801

More information

Data Analysis 3. Support Vector Machines. Jan Platoš October 30, 2017

Data Analysis 3. Support Vector Machines. Jan Platoš October 30, 2017 Data Analysis 3 Support Vector Machines Jan Platoš October 30, 2017 Department of Computer Science Faculty of Electrical Engineering and Computer Science VŠB - Technical University of Ostrava Table of

More information

Independent components of face images: A represent at ion for face recognit ion

Independent components of face images: A represent at ion for face recognit ion 4th Joint Symposium on Neural Computation Proceedings 1997 Independent components of face images: A represent at ion for face recognit ion Marian Stewart Bartlett Terrence J. Sejnowski University of California

More information

Enhancing Performance of Face Recognition System Using Independent Component Analysis

Enhancing Performance of Face Recognition System Using Independent Component Analysis Enhancing Performance of Face Recognition System Using Independent Component Analysis Dipti Rane 1, Prof. Uday Bhave 2, and Asst Prof. Manimala Mahato 3 Student, Computer Science, Shah and Anchor Kuttchi

More information

Mobile Face Recognization

Mobile Face Recognization Mobile Face Recognization CS4670 Final Project Cooper Bills and Jason Yosinski {csb88,jy495}@cornell.edu December 12, 2010 Abstract We created a mobile based system for detecting faces within a picture

More information

Support Vector Machines

Support Vector Machines Support Vector Machines RBF-networks Support Vector Machines Good Decision Boundary Optimization Problem Soft margin Hyperplane Non-linear Decision Boundary Kernel-Trick Approximation Accurancy Overtraining

More information

The Analysis of Parameters t and k of LPP on Several Famous Face Databases

The Analysis of Parameters t and k of LPP on Several Famous Face Databases The Analysis of Parameters t and k of LPP on Several Famous Face Databases Sujing Wang, Na Zhang, Mingfang Sun, and Chunguang Zhou College of Computer Science and Technology, Jilin University, Changchun

More information

Support Vector Machines

Support Vector Machines Support Vector Machines RBF-networks Support Vector Machines Good Decision Boundary Optimization Problem Soft margin Hyperplane Non-linear Decision Boundary Kernel-Trick Approximation Accurancy Overtraining

More information

Genetic Feature Subset Selection for Gender Classification: A Comparison Study

Genetic Feature Subset Selection for Gender Classification: A Comparison Study Genetic Feature Subset Selection for Gender Classification: A Comparison Study Zehang Sun 1,George Bebis 1, Xiaojing Yuan 1, and Sushil J. Louis 2 1 Computer Vision Lab. Department of Computer Science,

More information

Face Recognition Using SIFT- PCA Feature Extraction and SVM Classifier

Face Recognition Using SIFT- PCA Feature Extraction and SVM Classifier IOSR Journal of VLSI and Signal Processing (IOSR-JVSP) Volume 5, Issue 2, Ver. II (Mar. - Apr. 2015), PP 31-35 e-issn: 2319 4200, p-issn No. : 2319 4197 www.iosrjournals.org Face Recognition Using SIFT-

More information

Face detection and recognition. Many slides adapted from K. Grauman and D. Lowe

Face detection and recognition. Many slides adapted from K. Grauman and D. Lowe Face detection and recognition Many slides adapted from K. Grauman and D. Lowe Face detection and recognition Detection Recognition Sally History Early face recognition systems: based on features and distances

More information

Unsupervised learning in Vision

Unsupervised learning in Vision Chapter 7 Unsupervised learning in Vision The fields of Computer Vision and Machine Learning complement each other in a very natural way: the aim of the former is to extract useful information from visual

More information

HW2 due on Thursday. Face Recognition: Dimensionality Reduction. Biometrics CSE 190 Lecture 11. Perceptron Revisited: Linear Separators

HW2 due on Thursday. Face Recognition: Dimensionality Reduction. Biometrics CSE 190 Lecture 11. Perceptron Revisited: Linear Separators HW due on Thursday Face Recognition: Dimensionality Reduction Biometrics CSE 190 Lecture 11 CSE190, Winter 010 CSE190, Winter 010 Perceptron Revisited: Linear Separators Binary classification can be viewed

More information

SVM Classification in Multiclass Letter Recognition System

SVM Classification in Multiclass Letter Recognition System Global Journal of Computer Science and Technology Software & Data Engineering Volume 13 Issue 9 Version 1.0 Year 2013 Type: Double Blind Peer Reviewed International Research Journal Publisher: Global Journals

More information

We are IntechOpen, the world s leading publisher of Open Access books Built by scientists, for scientists. International authors and editors

We are IntechOpen, the world s leading publisher of Open Access books Built by scientists, for scientists. International authors and editors We are IntechOpen, the world s leading publisher of Open Access books Built by scientists, for scientists 4,000 6,000 M Open access books available International authors and editors Downloads Our authors

More information

ViFaI: A trained video face indexing scheme

ViFaI: A trained video face indexing scheme ViFaI: A trained video face indexing scheme Harsh Nayyar hnayyar@stanford.edu Audrey Wei awei1001@stanford.edu Abstract In this work, we implement face identification of captured videos by first training

More information

Face Recognition by Combining Kernel Associative Memory and Gabor Transforms

Face Recognition by Combining Kernel Associative Memory and Gabor Transforms Face Recognition by Combining Kernel Associative Memory and Gabor Transforms Author Zhang, Bai-ling, Leung, Clement, Gao, Yongsheng Published 2006 Conference Title ICPR2006: 18th International Conference

More information

Face Recognition for Mobile Devices

Face Recognition for Mobile Devices Face Recognition for Mobile Devices Aditya Pabbaraju (adisrinu@umich.edu), Srujankumar Puchakayala (psrujan@umich.edu) INTRODUCTION Face recognition is an application used for identifying a person from

More information

Face Detection using Hierarchical SVM

Face Detection using Hierarchical SVM Face Detection using Hierarchical SVM ECE 795 Pattern Recognition Christos Kyrkou Fall Semester 2010 1. Introduction Face detection in video is the process of detecting and classifying small images extracted

More information

Support Vector Machines (a brief introduction) Adrian Bevan.

Support Vector Machines (a brief introduction) Adrian Bevan. Support Vector Machines (a brief introduction) Adrian Bevan email: a.j.bevan@qmul.ac.uk Outline! Overview:! Introduce the problem and review the various aspects that underpin the SVM concept.! Hard margin

More information

An Automatic Face Recognition System in the Near Infrared Spectrum

An Automatic Face Recognition System in the Near Infrared Spectrum An Automatic Face Recognition System in the Near Infrared Spectrum Shuyan Zhao and Rolf-Rainer Grigat Technical University Hamburg Harburg Vision Systems, 4-08/1 Harburger Schloßstr 20 21079 Hamburg, Germany

More information

A Different Approach to Appearance based Statistical Method for Face Recognition Using Median

A Different Approach to Appearance based Statistical Method for Face Recognition Using Median 262 IJCSNS International Journal of Computer Science and Network Security, VOL.7 No.4, April 2007 A Different Approach to Appearance based Statistical Method for Face Recognition Using Median M A Rabbani

More information

The Curse of Dimensionality

The Curse of Dimensionality The Curse of Dimensionality ACAS 2002 p1/66 Curse of Dimensionality The basic idea of the curse of dimensionality is that high dimensional data is difficult to work with for several reasons: Adding more

More information

Data Mining Final Project Francisco R. Ortega Professor: Dr. Tao Li

Data Mining Final Project Francisco R. Ortega Professor: Dr. Tao Li Data Mining Final Project Francisco R. Ortega Professor: Dr. Tao Li FALL 2009 1.Introduction In the data mining class one of the aspects of interest were classifications. For the final project, the decision

More information

Gender Recognition from Model s Face Using SVM Algorithm

Gender Recognition from Model s Face Using SVM Algorithm Gender Recognition from Model s Face Using SVM Algorithm Deepak Deshmukh Lecturer EXTC Dept. VOGCOE Dist. Thane. Mumbai. India Abstract - This paper presents a method for Gender Recognition using Support

More information

Discriminate Analysis

Discriminate Analysis Discriminate Analysis Outline Introduction Linear Discriminant Analysis Examples 1 Introduction What is Discriminant Analysis? Statistical technique to classify objects into mutually exclusive and exhaustive

More information

Diagonal Principal Component Analysis for Face Recognition

Diagonal Principal Component Analysis for Face Recognition Diagonal Principal Component nalysis for Face Recognition Daoqiang Zhang,2, Zhi-Hua Zhou * and Songcan Chen 2 National Laboratory for Novel Software echnology Nanjing University, Nanjing 20093, China 2

More information

Non-Bayesian Classifiers Part II: Linear Discriminants and Support Vector Machines

Non-Bayesian Classifiers Part II: Linear Discriminants and Support Vector Machines Non-Bayesian Classifiers Part II: Linear Discriminants and Support Vector Machines Selim Aksoy Department of Computer Engineering Bilkent University saksoy@cs.bilkent.edu.tr CS 551, Spring 2007 c 2007,

More information

Performance Analysis of PCA and LDA

Performance Analysis of PCA and LDA International Journal of Innovative Research in Electronics and Communications (IJIREC) Volume 2, Issue 2, April 2015, PP 17-22 ISSN 2349-4042 (Print) & ISSN 2349-4050 (Online) www.arcjournals.org Performance

More information

Feature scaling in support vector data description

Feature scaling in support vector data description Feature scaling in support vector data description P. Juszczak, D.M.J. Tax, R.P.W. Duin Pattern Recognition Group, Department of Applied Physics, Faculty of Applied Sciences, Delft University of Technology,

More information

Facial expression recognition using shape and texture information

Facial expression recognition using shape and texture information 1 Facial expression recognition using shape and texture information I. Kotsia 1 and I. Pitas 1 Aristotle University of Thessaloniki pitas@aiia.csd.auth.gr Department of Informatics Box 451 54124 Thessaloniki,

More information

Dr. Prakash B. Khanale 3 Dnyanopasak College, Parbhani, (M.S.), India

Dr. Prakash B. Khanale 3 Dnyanopasak College, Parbhani, (M.S.), India ISSN: 2321-7782 (Online) Volume 3, Issue 9, September 2015 International Journal of Advance Research in Computer Science and Management Studies Research Article / Survey Paper / Case Study Available online

More information

Face Recognition based on Discrete Cosine Transform and Support Vector Machines

Face Recognition based on Discrete Cosine Transform and Support Vector Machines Face Recognition based on Discrete Cosine Transform and Support Vector Machines GABRIEL CHAVES AFONSO COUTINHO CRISTIANO LEITE DE CASTRO 2 UFLA - Federal University of Lavras DCC - Department of Computer

More information

Feature Extraction for Nearest Neighbor Classification: Application to Gender Recognition

Feature Extraction for Nearest Neighbor Classification: Application to Gender Recognition Feature Extraction for Nearest Neighbor Classification: Application to Gender Recognition David Masip,* Jordi Vitrià Centre de Visió per Computador (CVC), Departemente Informàtica, Universitat Autònoma

More information

Topic 4: Support Vector Machines

Topic 4: Support Vector Machines CS 4850/6850: Introduction to achine Learning Fall 2018 Topic 4: Support Vector achines Instructor: Daniel L Pimentel-Alarcón c Copyright 2018 41 Introduction Support vector machines (SVs) are considered

More information

Binarization of Color Character Strings in Scene Images Using K-means Clustering and Support Vector Machines

Binarization of Color Character Strings in Scene Images Using K-means Clustering and Support Vector Machines 2011 International Conference on Document Analysis and Recognition Binarization of Color Character Strings in Scene Images Using K-means Clustering and Support Vector Machines Toru Wakahara Kohei Kita

More information

Laplacian MinMax Discriminant Projection and its Applications

Laplacian MinMax Discriminant Projection and its Applications Laplacian MinMax Discriminant Projection and its Applications Zhonglong Zheng and Xueping Chang Department of Computer Science, Zhejiang Normal University, Jinhua, China Email: zhonglong@sjtu.org Jie Yang

More information

Automated Canvas Analysis for Painting Conservation. By Brendan Tobin

Automated Canvas Analysis for Painting Conservation. By Brendan Tobin Automated Canvas Analysis for Painting Conservation By Brendan Tobin 1. Motivation Distinctive variations in the spacings between threads in a painting's canvas can be used to show that two sections of

More information

Face Recognition using Eigenfaces SMAI Course Project

Face Recognition using Eigenfaces SMAI Course Project Face Recognition using Eigenfaces SMAI Course Project Satarupa Guha IIIT Hyderabad 201307566 satarupa.guha@research.iiit.ac.in Ayushi Dalmia IIIT Hyderabad 201307565 ayushi.dalmia@research.iiit.ac.in Abstract

More information

Restricted Nearest Feature Line with Ellipse for Face Recognition

Restricted Nearest Feature Line with Ellipse for Face Recognition Journal of Information Hiding and Multimedia Signal Processing c 2012 ISSN 2073-4212 Ubiquitous International Volume 3, Number 3, July 2012 Restricted Nearest Feature Line with Ellipse for Face Recognition

More information

Class-Information-Incorporated Principal Component Analysis

Class-Information-Incorporated Principal Component Analysis Class-Information-Incorporated Principal Component Analysis Songcan Chen * Tingkai Sun Dept. of Computer Science & Engineering, Nanjing University of Aeronautics & Astronautics, Nanjing, 210016, China

More information

International Journal of Digital Application & Contemporary research Website: (Volume 1, Issue 8, March 2013)

International Journal of Digital Application & Contemporary research Website:   (Volume 1, Issue 8, March 2013) Face Recognition using ICA for Biometric Security System Meenakshi A.D. Abstract An amount of current face recognition procedures use face representations originate by unsupervised statistical approaches.

More information

A GENERIC FACE REPRESENTATION APPROACH FOR LOCAL APPEARANCE BASED FACE VERIFICATION

A GENERIC FACE REPRESENTATION APPROACH FOR LOCAL APPEARANCE BASED FACE VERIFICATION A GENERIC FACE REPRESENTATION APPROACH FOR LOCAL APPEARANCE BASED FACE VERIFICATION Hazim Kemal Ekenel, Rainer Stiefelhagen Interactive Systems Labs, Universität Karlsruhe (TH) 76131 Karlsruhe, Germany

More information

Face detection and recognition. Detection Recognition Sally

Face detection and recognition. Detection Recognition Sally Face detection and recognition Detection Recognition Sally Face detection & recognition Viola & Jones detector Available in open CV Face recognition Eigenfaces for face recognition Metric learning identification

More information

Extended Isomap for Pattern Classification

Extended Isomap for Pattern Classification From: AAAI- Proceedings. Copyright, AAAI (www.aaai.org). All rights reserved. Extended for Pattern Classification Ming-Hsuan Yang Honda Fundamental Research Labs Mountain View, CA 944 myang@hra.com Abstract

More information

Equation to LaTeX. Abhinav Rastogi, Sevy Harris. I. Introduction. Segmentation.

Equation to LaTeX. Abhinav Rastogi, Sevy Harris. I. Introduction. Segmentation. Equation to LaTeX Abhinav Rastogi, Sevy Harris {arastogi,sharris5}@stanford.edu I. Introduction Copying equations from a pdf file to a LaTeX document can be time consuming because there is no easy way

More information

Face and Facial Expression Detection Using Viola-Jones and PCA Algorithm

Face and Facial Expression Detection Using Viola-Jones and PCA Algorithm Face and Facial Expression Detection Using Viola-Jones and PCA Algorithm MandaVema Reddy M.Tech (Computer Science) Mailmv999@gmail.com Abstract Facial expression is a prominent posture beneath the skin

More information

Bagging for One-Class Learning

Bagging for One-Class Learning Bagging for One-Class Learning David Kamm December 13, 2008 1 Introduction Consider the following outlier detection problem: suppose you are given an unlabeled data set and make the assumptions that one

More information

The Effects of Outliers on Support Vector Machines

The Effects of Outliers on Support Vector Machines The Effects of Outliers on Support Vector Machines Josh Hoak jrhoak@gmail.com Portland State University Abstract. Many techniques have been developed for mitigating the effects of outliers on the results

More information

Multi-Pose Face Recognition And Tracking System

Multi-Pose Face Recognition And Tracking System Available online at www.sciencedirect.com Procedia Computer Science 6 (2011) 381 386 Complex Adaptive Systems, Volume 1 Cihan H. Dagli, Editor in Chief Conference Organized by Missouri University of Science

More information

LOCAL APPEARANCE BASED FACE RECOGNITION USING DISCRETE COSINE TRANSFORM

LOCAL APPEARANCE BASED FACE RECOGNITION USING DISCRETE COSINE TRANSFORM LOCAL APPEARANCE BASED FACE RECOGNITION USING DISCRETE COSINE TRANSFORM Hazim Kemal Ekenel, Rainer Stiefelhagen Interactive Systems Labs, University of Karlsruhe Am Fasanengarten 5, 76131, Karlsruhe, Germany

More information

Verification: is that a lamp? What do we mean by recognition? Recognition. Recognition

Verification: is that a lamp? What do we mean by recognition? Recognition. Recognition Recognition Recognition The Margaret Thatcher Illusion, by Peter Thompson The Margaret Thatcher Illusion, by Peter Thompson Readings C. Bishop, Neural Networks for Pattern Recognition, Oxford University

More information

FACE RECOGNITION UNDER LOSSY COMPRESSION. Mustafa Ersel Kamaşak and Bülent Sankur

FACE RECOGNITION UNDER LOSSY COMPRESSION. Mustafa Ersel Kamaşak and Bülent Sankur FACE RECOGNITION UNDER LOSSY COMPRESSION Mustafa Ersel Kamaşak and Bülent Sankur Signal and Image Processing Laboratory (BUSIM) Department of Electrical and Electronics Engineering Boǧaziçi University

More information

Published by: PIONEER RESEARCH & DEVELOPMENT GROUP(www.prdg.org) 1

Published by: PIONEER RESEARCH & DEVELOPMENT GROUP(www.prdg.org) 1 FACE RECOGNITION USING PRINCIPLE COMPONENT ANALYSIS (PCA) ALGORITHM P.Priyanka 1, Dorairaj Sukanya 2 and V.Sumathy 3 1,2,3 Department of Computer Science and Engineering, Kingston Engineering College,

More information

Fraud Detection using Machine Learning

Fraud Detection using Machine Learning Fraud Detection using Machine Learning Aditya Oza - aditya19@stanford.edu Abstract Recent research has shown that machine learning techniques have been applied very effectively to the problem of payments

More information

Implementation of a Face Detection and Recognition System

Implementation of a Face Detection and Recognition System Implementation of a Face Detection and Recognition System Sanghyun Choi Johns Hopkins University 3400 N. Charles Street Baltimore, MD 21218, USA schoi60@jhu.edu Michael Flynn Johns Hopkins University 3400

More information