A Real-time Application of Hand Gesture Recognition in Video Games

Size: px
Start display at page:

Download "A Real-time Application of Hand Gesture Recognition in Video Games"

Transcription

1 A Real-time Application of Hand Gesture Recognition in Video Games Ying Li March 10, 2011 Abstract Hand gesture recognition has many useful applications, and one of those is video game. In this project, a virtual Rubik s cube game is implemented, in which the cube is manipulated by recognizing the hand gesture of the user. The application uses skin color to detect the hands and extracts features which contain spatiotemporal information to be classified by support vector machine (SVM). In the experiment, the application achieves a recognition rate higher than 90%, and each frame is processed and rendered in 30ms. 1. Introduction With the rapid development of video games and 3D engines, the role of human computer interaction (HCI), is becoming more important. Tremendous efforts have been made to provide a more natural and realistic experience interacting with the computer. In contrast to a video renderer, which projects a 3D world into 2D images, conventional input devices, such as mice and joysticks, convert 2D motion to manipulate the 3D objects in the games, though human hand gesture and body language, on the other hand, offer a more natural way of direct manipulation. Out of this consideration, this project explores the possibility of real-time recognition of hand gestures in a simple cube game. Approaches of hand gesture recognition can mainly be classified into two categories: those trying to infer gestures directly from the visual images observed, usually denoted as appearance-based modeling, and those inferring gestures from the motion and posture model parameters computed from the image sequences, known as 3D-model-based[1, 2]. Though 3D-model-based approaches offer a general solution to most of the hand gesture recognition problems, it is computationally intensive and the intermediate computation results might be more than necessary for a specific task. Moreover the transformation from 3D models to 2D images might not be invertible, which makes the 3D-model-based approaches unstable in some cases. In contrast, the appearance-based approach can be realized in real-time and achieve high accuracy for specific tasks. The recognition process in appearance-based approaches consists of the following steps: the spotting of the hand blob, extraction of features, and classification, depicted in fig.1. Drawing upon the similarity to the speech recognition problem, Starner et al proposed using HMM to recognize sentence-level American Sign Language (ASL) selected from a 40-word lexicon, achieving an accuracy higher than 90%[3], whereas Yang et al reported 93.42% recognition rate by classifying the motion trajectories using time-delay neural network (TDNN)[4]. Darrell et al took a different approach by accumulating representative n-tuple image sequences as gesture templates[5]. Another approach exploiting the image sequences is to model different gestural actions by motion history images (MHI), which are 2D images formed by accumulating the motion of every single pixel in the image over some temporal window[7]. In this project however, an SVM is used for the recognition task, given their success in image classification problems. Figure 1: major steps in hand gesture recognition 2. Approach As depicted in fig.1, the feature vector is extracted from the hand blobs separated from the visual image background, which is then fed to the SVM classifier. Since an SVM does not use temporal information directly, we extract features that encode spatiotemperal data. 1

2 2.1. Detection of hand blob Provided that the user can interact with the application, the algorithm simply uses a skin color model determined a priori to segment the hands. Anyhow such a skin color model might not be appropriate when there are abrupt changes in lighting conditions or the difference between the hands and the background is not obvious. Nevertheless the segmentation problem in this case is moderately alleviated because the field of view of a webcam is limited and the interest regions are restricted around the virtual cube. After the mask of potential hand pixels is computed, a morphological open and close operation is applied to the mask to remove noises and join disparate elements. Then a breadth-first-search (BFS) is implemented upon the modified mask to find connected hand components. The centroid and bounding box of the component is calculated as a by-product of the process Feature extraction After the hands are segmented from the background, a second analysis is applied to each hand blob detected to form the feature vector. Along with the centroid x, y coordinates mentioned in hand blob detection section, changes in x, y between the frames, optical flow and shape information are included also. Each hand component is labeled with right, top, left or bottom according to its relative position to the center of the virtual cube Optical flow computation To compute the optical flow between the frames, Lucas- Kanade optical flow method is used upon every two consecutive grayscale images. A brief explanation of the algorithm is presented below. When calculating optical flow from one frame to the next, it is assumed that the brightness of the same point remains constant: I(x, y, t) =I(x + δx, y + δy, t + δt) x, y are the coordinates of a pixel, and t is the time when the first frame is taken. Applying Taylor series expansion to the RHS of the equation, we get: I(x+δx, y+δy, t+δt) =I(x, y, t)+ δx+ δy+ δx δy δt δt + H.O.T. From this equation it follows that: δx δx + δy δy + δt =0 δt Dividing by δt we get: which results in δx δx δt + δy δy δt + δt δt δt =0 δx V x + δy V y + δt =0 where V x,v y are the x and y components of the velocity and δx, δy and δt are the derivatives of the image at (x, y, t) in the corresponding directions. Lucas and Kanade assume a translational model and solve for a velocity that approximately satisfies the equation above for all the pixels in a small neighborhood of size N N. Inthisway,weobtainanoverconstrained linear system of equations, of which the solution can be solved by least square method: in which I(x, y, t) = and the quantity A = b = I(p 1 ) T I(p 2 ) T.. I(p N 2) T I t (p 1 ) I t (p 2 ). I t (p N 2) [ Ix (x, y, t) I y (x, y, t) ] = [ δx δy I t (x, y, t) = (x, y, t) δt ] (x, y, t) (x, y, t) that the solution can be written as: (A T A) 1 A T b. To expedite the speed, the computation of the optical flow is refined to the region deduced from hand detection section as the bounding box. The x, y velocities calculated for each pixel are then assigned to an 8x8 grid to form a 128-element vector. To put more emphasis upon the contours, another 128-element vector is constructed from the optical flow of contours. Though computing optical flow of the interest points will conduce to the confidence of the accuracy, it is adversely proposed in this case, since the hand blob is blurring in the action phase, making the corner detection inaccurate. 2

3 Hand shape related features Geometric moments are succinct description of the shape of a component; therefore the second moments (m 20, m 11, m 02 ) are computed from the contour of each hand: [ m20 m 11 m 11 m 02 ] By curve fitting, the major and minor axes of the ellipse are deduced as features. To maintain more detailed shape information of the hand, each pixel in the hand component is assigned to an 8x8 grid and the sum of x, y coordinates of each grid is retrieved to form a 128-element vector. With reference to the duality property of SVM, the outer product is not computed for curve fitting, neither are the eigenvalues or eigenvectors, to mitigate computation cost. As it is in optical flow computation, another 128-element vector is constructed from the contour of the hand Feature vector normalization Because the features extracted in the previous section are of different orders of magnitude, which make the average and standard deviation between different features incomparable, the features are therefore normalized. For the tf-idf matrix computed from the training set, the average and standard deviation is calculated for every feature, and then each element is divided by standard deviation after subtracting the average. If the standard deviation of a feature equals to zero, then that column of feature is removed from the tf-idf matrix. This process imposes a Gaussian distribution upon every feature; every feature has average equaling zero and standard deviation equaling one. To transform all the elements to the range [-1,1], each element is divided by the square root of the number of training samples, noticing that: (x i x) 2 N (x i x) 2 x i x N (x i x) 2 = σ N Though SVM might be more immune to unnormalized dataset compared with other method by learning weight for different features in the training process, a normalization step would possibly conduce to the accuracy of the computation because the storage of a number is limited to its type size Brief explanation of SVM The normalized feature vector structured in the form described above is then passed to the SVM classifier. Unlike conventional classifiers, SVM does not train to fit the sample density of distribution; instead it trains to minimize structural risk[6]. A brief explanation of SVM is described as below. Figure 2: illustration of linear SVM Given dataset x i,y i, with y i { 1, +1}, x i R n, if there exists a hyperplane which can separate the dataset, it is written as: g(x) =wx+b =0. H 1 is a hyperplane parallel to H consisting of the closest points labeled with 1, whereas H 2 is a hyperplane parallel to H consisting of the closest points labeled with +1, so we get: H : wx+ b =0; H 1 : wx+ b = k 1 ; H 2 : wx+ b = k 2 Assuming k = k1 k2 2,then H 1 : wx + b k 1 + k = k; which can be also be written as: H 1 : wx + b = k; H 2 : wx + b k 2 k = k H 2 : wx + b = k Deviding both equations with k, we have: H 1 : wx + b =1; H 2 : wx + b = 1 From the equations above, it can be derived that the 2 margin is w. So to maximize the margin, we compute min 1 2 w 2 s.t. y i (wx i + b) 1(i =1, 2,...,n) Because all the functions are convex, there exists a unique global optimum. Solving the problem using Lagrange multiplier α i,weget: w = y i α i x i,b= y j y i α i (x i x j ), j {j α j > 0} 3

4 To allow for mislabeled samples, a slack variable ξ is introduced to trade off between the large margin and a small error penalty: y i (wx i + b) 1 ξ i Now to maximize the margin, we compute min( 1 2 w 2 + C ξ i ) y i (wx i + b) 1 ξ i (i =1, 2,...,n) ξ i 0 The classifying function we thus derived is: f(x) =sgn{wx + b} =sgn{ α i y i (x i x)+b} Referring to the equation above, it follows that to compute the label of a test sample, we just need to compute the inner products of the test sample and support vectors. If the dataset is not linearly separable in the input space, the samples can first be projected to a higher space and then be classified with the linear separation method outlined above, and that is where the kernel trick comes in. f(x) =sgn{ l α i y i K(x i,x)+b} K(x i,y i )=(φ(x i ) φ(y i )) If K satisfies Mercer s Conditions, then the kernel represents a legitimate inner product in feature space. This strategy is of resemblance to the mechanism of neural network, with the hidden layer being the inner products of the test sample and support vectors (fig.3). Therefore, the choice of kernel is another parameter to SVM method, though it is reported that the performance remains complaisant whatever the choice of kernel is in some applications[6]. In this project, Gaussian or Radial Basis Function (RBF) is used because of its effectiveness in image recognitions: K(x i,x j )=exp( γ x i x j 2 ), for γ>0. From the optimizing method outlined above, that SVM is minimizing the weight w 2 naturally reduces the chance of overfitting. Moreover its structural risk minimizing principle makes it of advantage when applied to a dataset of very high dimension. Though the prediction of an SVM classifier is only determined by the support vectors, in this project however, it is difficult to figure out which are the samples on the margin and therefore construct the training set. Figure 3: illustration of the kernel trick 2.5. Game status update rule The probabilities of p(actiona actionb), p(action inaction) and p(inaction action) can be derived by applying the classifier to the test set. If a window of specific temporal length (less a typical action) is set, the action update rule can be theoretically deduced by maximizing the likelihood of the multinomial distribution. Anyhow with consideration that the sample set is rather limited compared with the untested user actions, it is likely that such delicate strategies might not work as well as simple strategies. So in the application, the update rule is tentatively set that if the predicted labels of two consecutive frames are the same, the status of the cube will be updated if the current status is not the same as the predicted one. 3. Experiment Results and Analysis The dataset is built by recording frames at 25 fps from a webcam. True labels of the dataset are determined afterwards. Anyhow because of the subtlety when a previous action is retracted whereas the next one is in preparation, there is no guarantee that the current labeling is correct. The state of the frame is classified into 49 groups, in which the rotations for preparation purposes are isolated from the null state. If two hands are detected in the frame at appropriate positions, with one hand rotating and the other keeping still, the corresponding level of the cube is twisted. Therefore, the middle level of the cube cannot be manipulated in any case. Anyhow this is not a problem because the viewpoint of the cube can be changed. If only a single hand is detected 4

5 and the rotation is correctly perceived by the application, the corresponding rotation of the cube will take place. Reducing the number of states is likely to reduce the computation cost, especially if the SVM classifier is not constructed as decision tree. However, it is unlikely that the reduction of states would contribute to the decrease of error rate, which depends upon the construction of support vectors. The reason why the nonrotation actions are separated from the null state is that those labels might be of use to construct a finite state machine to update the status of the cube if the prediction of each frame is adequately accurate. a b Figure 4: a. the gesture of rotating the cube upward; b. the gesture of rotating the cube downward From fig.4, the optical flow labeled upon the visual images is in accordance with the movement of the hand. Those frames were sampled at 25 fps. Since Lucas and Kanade method imposes a translational model upon the transformation, increasing the sampling rate might be of help to the accuracy of optical flow calculation, provided that most of the reactions involved in this application are not pure translations, and might possibly be accompanied with shape deformations. The accuracy of the SVM classifier is outlined in table.1 as in a grid search for the appropriate parameters by 10-fold cross validation. The best recognition rate achieved is 91.87% when C equals to 32 and gamma equals to 2. It is shown in the figure that as gamma increases, the accuracy first increases accordingly and then drops down. This behavior is correlated to the influence of gamma in the Gaussian function. As gamma becomes bigger, the Gaussian function is steeper. When accompanied with a larger penalty (C value), the SVM classifier is able to reproduce highly irregular decision boundaries, at the risk of overfitting to the training set. ThesamepatternshowninthetablewhenCincreases also demonstrate this. From the table, most of the errors come from mistaking hand actions for inactions, which means the application might be inert to the action of users. On the other hand, the errors coming from mistaking inaction for actions would result in unintended movement of the cube when it is supposed to be still. Anyhow the low rate of taking one action to another implies the potential of using an SVM classifier in hand gesture recognition. Given the probability of p(actiona actionb), p(action inaction) and p(inaction action), it can be derived that using the current cube update rule, the probability of correct status change is if the previous status is inactive and the current status is active, and for the reversed status change. Anyhow the statistics calculated above is based upon the result from cross validation, and the performance of the application might be mitigated in practice. Examining the error labels, most of the mistakes occur at the boundary of two actions. Because of the ambiguity between the preparation and the retraction phases of continuous gestures, mistakes of this type are difficult to ameliorate. I expect that around 10% of the mistakes might be corrected if the trajectory of the gesture, or at least the state of the previous frame is known. When tested on a 2.53GHz computer, the application extracts features in 15ms, predicting the label of a frame by SVM classifier in 12ms, and does the rendering in 3ms; all image operation functions and the implementation of SVM are from OpenCV lib. 4. Conclusion In the project, a real-time application of hand gesture recognition is implemented. The application uses skin color to detect hand blobs and extracts features from motion and shape of the hands as input to an SVM classifier, which outputs a label. Though the error rate 5

6 [2] T. S Huang, and V. Pavlovic, Hand gesture modeling, analysis, and synthesis, Proc IEEE International Workshop on Automatic Face and Gesture Recognition, September 1995, pp [3] T. Starner, J. Weaver, and A. Pentland, Real-time American Sign Language recognition using desk and wearable computer based video, IEEE Trans. Pattern Anal. Mach. Intell. 20, 1998, [4] M-H. Yang, N. Ahuja, Recognizing hand gesture using motion trajectories. Proc. of IEEE CS Conference on Computer Vision and Pattern Recognition. 1999, pp [5] T. Darrell, I. Essa, and A. Pentland, Task-Specific Gesture Analysis in Real-Time Using nterpolated Views, IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 18, no. 12, pp.,236-1,242, Dec [6] N. Cristianini and J. Shawe-Taylor, An Introduction to Support Vector Machines and Other Kernel-based Learning Methods, Cambridge University Press, [7] A.F. Bobick and J.W. Davis, Real-Time Recognition of ActivityUsing Temporal Templates, Proc. Int l Conf. Automatic Face and Gesture Recognition, Killington, Vt., Oct Table 1: Error rate computed from 10-fold cross validation in a grid search for appropriate gamma and C is , it can be further reduced by taking the trajectory into consideration. The error rate is computed from a dataset in which the samples are collected at a rather consistent position relative to the camera, so it is not predictable how the system will behave when the hand position changes even if the training set is enlarged. Nor is it certain how the application will react to different people. References [1] V. Pavlovic, R. Sharma, and T. S. Huang, Visual interpretation of hand gestures for human-computer interaction: A review, IEEE Trans. Pattern Anal. Mach. Intell. 19, 1997,

A Two-stage Scheme for Dynamic Hand Gesture Recognition

A Two-stage Scheme for Dynamic Hand Gesture Recognition A Two-stage Scheme for Dynamic Hand Gesture Recognition James P. Mammen, Subhasis Chaudhuri and Tushar Agrawal (james,sc,tush)@ee.iitb.ac.in Department of Electrical Engg. Indian Institute of Technology,

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

Matching. Compare region of image to region of image. Today, simplest kind of matching. Intensities similar.

Matching. Compare region of image to region of image. Today, simplest kind of matching. Intensities similar. Matching Compare region of image to region of image. We talked about this for stereo. Important for motion. Epipolar constraint unknown. But motion small. Recognition Find object in image. Recognize object.

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

CS201: Computer Vision Introduction to Tracking

CS201: Computer Vision Introduction to Tracking CS201: Computer Vision Introduction to Tracking John Magee 18 November 2014 Slides courtesy of: Diane H. Theriault Question of the Day How can we represent and use motion in images? 1 What is Motion? Change

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

HUMAN COMPUTER INTERFACE BASED ON HAND TRACKING

HUMAN COMPUTER INTERFACE BASED ON HAND TRACKING Proceedings of MUSME 2011, the International Symposium on Multibody Systems and Mechatronics Valencia, Spain, 25-28 October 2011 HUMAN COMPUTER INTERFACE BASED ON HAND TRACKING Pedro Achanccaray, Cristian

More information

Bagging and Boosting Algorithms for Support Vector Machine Classifiers

Bagging and Boosting Algorithms for Support Vector Machine Classifiers Bagging and Boosting Algorithms for Support Vector Machine Classifiers Noritaka SHIGEI and Hiromi MIYAJIMA Dept. of Electrical and Electronics Engineering, Kagoshima University 1-21-40, Korimoto, Kagoshima

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

SUPPORT VECTOR MACHINES

SUPPORT VECTOR MACHINES SUPPORT VECTOR MACHINES Today Reading AIMA 18.9 Goals (Naïve Bayes classifiers) Support vector machines 1 Support Vector Machines (SVMs) SVMs are probably the most popular off-the-shelf classifier! Software

More information

Chap.12 Kernel methods [Book, Chap.7]

Chap.12 Kernel methods [Book, Chap.7] Chap.12 Kernel methods [Book, Chap.7] Neural network methods became popular in the mid to late 1980s, but by the mid to late 1990s, kernel methods have also become popular in machine learning. The first

More information

Data Mining: Concepts and Techniques. Chapter 9 Classification: Support Vector Machines. Support Vector Machines (SVMs)

Data Mining: Concepts and Techniques. Chapter 9 Classification: Support Vector Machines. Support Vector Machines (SVMs) Data Mining: Concepts and Techniques Chapter 9 Classification: Support Vector Machines 1 Support Vector Machines (SVMs) SVMs are a set of related supervised learning methods used for classification Based

More information

COMPUTER VISION > OPTICAL FLOW UTRECHT UNIVERSITY RONALD POPPE

COMPUTER VISION > OPTICAL FLOW UTRECHT UNIVERSITY RONALD POPPE COMPUTER VISION 2017-2018 > OPTICAL FLOW UTRECHT UNIVERSITY RONALD POPPE OUTLINE Optical flow Lucas-Kanade Horn-Schunck Applications of optical flow Optical flow tracking Histograms of oriented flow Assignment

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

Augmented Reality VU. Computer Vision 3D Registration (2) Prof. Vincent Lepetit

Augmented Reality VU. Computer Vision 3D Registration (2) Prof. Vincent Lepetit Augmented Reality VU Computer Vision 3D Registration (2) Prof. Vincent Lepetit Feature Point-Based 3D Tracking Feature Points for 3D Tracking Much less ambiguous than edges; Point-to-point reprojection

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

Robert Collins CSE598G. Intro to Template Matching and the Lucas-Kanade Method

Robert Collins CSE598G. Intro to Template Matching and the Lucas-Kanade Method Intro to Template Matching and the Lucas-Kanade Method Appearance-Based Tracking current frame + previous location likelihood over object location current location appearance model (e.g. image template,

More information

A Short SVM (Support Vector Machine) Tutorial

A Short SVM (Support Vector Machine) Tutorial A Short SVM (Support Vector Machine) Tutorial j.p.lewis CGIT Lab / IMSC U. Southern California version 0.zz dec 004 This tutorial assumes you are familiar with linear algebra and equality-constrained optimization/lagrange

More information

LOGISTIC REGRESSION FOR MULTIPLE CLASSES

LOGISTIC REGRESSION FOR MULTIPLE CLASSES Peter Orbanz Applied Data Mining Not examinable. 111 LOGISTIC REGRESSION FOR MULTIPLE CLASSES Bernoulli and multinomial distributions The mulitnomial distribution of N draws from K categories with parameter

More information

Introduction to Support Vector Machines

Introduction to Support Vector Machines Introduction to Support Vector Machines CS 536: Machine Learning Littman (Wu, TA) Administration Slides borrowed from Martin Law (from the web). 1 Outline History of support vector machines (SVM) Two classes,

More information

A Performance Evaluation of HMM and DTW for Gesture Recognition

A Performance Evaluation of HMM and DTW for Gesture Recognition A Performance Evaluation of HMM and DTW for Gesture Recognition Josep Maria Carmona and Joan Climent Barcelona Tech (UPC), Spain Abstract. It is unclear whether Hidden Markov Models (HMMs) or Dynamic Time

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

SUPPORT VECTOR MACHINES

SUPPORT VECTOR MACHINES SUPPORT VECTOR MACHINES Today Reading AIMA 8.9 (SVMs) Goals Finish Backpropagation Support vector machines Backpropagation. Begin with randomly initialized weights 2. Apply the neural network to each training

More information

Scale-Invariance of Support Vector Machines based on the Triangular Kernel. Abstract

Scale-Invariance of Support Vector Machines based on the Triangular Kernel. Abstract Scale-Invariance of Support Vector Machines based on the Triangular Kernel François Fleuret Hichem Sahbi IMEDIA Research Group INRIA Domaine de Voluceau 78150 Le Chesnay, France Abstract This paper focuses

More information

Robust Lip Contour Extraction using Separability of Multi-Dimensional Distributions

Robust Lip Contour Extraction using Separability of Multi-Dimensional Distributions Robust Lip Contour Extraction using Separability of Multi-Dimensional Distributions Tomokazu Wakasugi, Masahide Nishiura and Kazuhiro Fukui Corporate Research and Development Center, Toshiba Corporation

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

Static Gesture Recognition with Restricted Boltzmann Machines

Static Gesture Recognition with Restricted Boltzmann Machines Static Gesture Recognition with Restricted Boltzmann Machines Peter O Donovan Department of Computer Science, University of Toronto 6 Kings College Rd, M5S 3G4, Canada odonovan@dgp.toronto.edu Abstract

More information

Towards the completion of assignment 1

Towards the completion of assignment 1 Towards the completion of assignment 1 What to do for calibration What to do for point matching What to do for tracking What to do for GUI COMPSCI 773 Feature Point Detection Why study feature point detection?

More information

Table of Contents. Recognition of Facial Gestures... 1 Attila Fazekas

Table of Contents. Recognition of Facial Gestures... 1 Attila Fazekas Table of Contents Recognition of Facial Gestures...................................... 1 Attila Fazekas II Recognition of Facial Gestures Attila Fazekas University of Debrecen, Institute of Informatics

More information

CS 4495 Computer Vision Motion and Optic Flow

CS 4495 Computer Vision Motion and Optic Flow CS 4495 Computer Vision Aaron Bobick School of Interactive Computing Administrivia PS4 is out, due Sunday Oct 27 th. All relevant lectures posted Details about Problem Set: You may *not* use built in Harris

More information

Moving Object Tracking in Video Using MATLAB

Moving Object Tracking in Video Using MATLAB Moving Object Tracking in Video Using MATLAB Bhavana C. Bendale, Prof. Anil R. Karwankar Abstract In this paper a method is described for tracking moving objects from a sequence of video frame. This method

More information

Chapter 9 Object Tracking an Overview

Chapter 9 Object Tracking an Overview Chapter 9 Object Tracking an Overview The output of the background subtraction algorithm, described in the previous chapter, is a classification (segmentation) of pixels into foreground pixels (those belonging

More information

Real-time Adaptive Hand Motion Recognition using a Sparse Bayesian Classifier

Real-time Adaptive Hand Motion Recognition using a Sparse Bayesian Classifier Real-time Adaptive Hand Motion Recognition using a Sparse Bayesian Classifier Shu-Fai Wong and Roberto Cipolla Department of Engineering, The University of Cambridge {sfw26, cipolla}@cam.ac.uk Abstract.

More information

Gesture Recognition Using Image Comparison Methods

Gesture Recognition Using Image Comparison Methods Gesture Recognition Using Image Comparison Methods Philippe Dreuw, Daniel Keysers, Thomas Deselaers, and Hermann Ney Lehrstuhl für Informatik VI Computer Science Department, RWTH Aachen University D-52056

More information

Linear methods for supervised learning

Linear methods for supervised learning Linear methods for supervised learning LDA Logistic regression Naïve Bayes PLA Maximum margin hyperplanes Soft-margin hyperplanes Least squares resgression Ridge regression Nonlinear feature maps Sometimes

More information

Machine Learning: Think Big and Parallel

Machine Learning: Think Big and Parallel Day 1 Inderjit S. Dhillon Dept of Computer Science UT Austin CS395T: Topics in Multicore Programming Oct 1, 2013 Outline Scikit-learn: Machine Learning in Python Supervised Learning day1 Regression: Least

More information

Human Hand Gesture Recognition Using Motion Orientation Histogram for Interaction of Handicapped Persons with Computer

Human Hand Gesture Recognition Using Motion Orientation Histogram for Interaction of Handicapped Persons with Computer Human Hand Gesture Recognition Using Motion Orientation Histogram for Interaction of Handicapped Persons with Computer Maryam Vafadar and Alireza Behrad Faculty of Engineering, Shahed University Tehran,

More information

Face Detection and Recognition in an Image Sequence using Eigenedginess

Face Detection and Recognition in an Image Sequence using Eigenedginess Face Detection and Recognition in an Image Sequence using Eigenedginess B S Venkatesh, S Palanivel and B Yegnanarayana Department of Computer Science and Engineering. Indian Institute of Technology, Madras

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

Support Vector Machines

Support Vector Machines Support Vector Machines About the Name... A Support Vector A training sample used to define classification boundaries in SVMs located near class boundaries Support Vector Machines Binary classifiers whose

More information

Classification by Support Vector Machines

Classification by Support Vector Machines Classification by Support Vector Machines Florian Markowetz Max-Planck-Institute for Molecular Genetics Computational Molecular Biology Berlin Practical DNA Microarray Analysis 2003 1 Overview I II III

More information

EE795: Computer Vision and Intelligent Systems

EE795: Computer Vision and Intelligent Systems EE795: Computer Vision and Intelligent Systems Spring 2012 TTh 17:30-18:45 FDH 204 Lecture 14 130307 http://www.ee.unlv.edu/~b1morris/ecg795/ 2 Outline Review Stereo Dense Motion Estimation Translational

More information

Discriminative classifiers for image recognition

Discriminative classifiers for image recognition Discriminative classifiers for image recognition May 26 th, 2015 Yong Jae Lee UC Davis Outline Last time: window-based generic object detection basic pipeline face detection with boosting as case study

More information

Motion. 1 Introduction. 2 Optical Flow. Sohaib A Khan. 2.1 Brightness Constancy Equation

Motion. 1 Introduction. 2 Optical Flow. Sohaib A Khan. 2.1 Brightness Constancy Equation Motion Sohaib A Khan 1 Introduction So far, we have dealing with single images of a static scene taken by a fixed camera. Here we will deal with sequence of images taken at different time intervals. Motion

More information

Capturing, Modeling, Rendering 3D Structures

Capturing, Modeling, Rendering 3D Structures Computer Vision Approach Capturing, Modeling, Rendering 3D Structures Calculate pixel correspondences and extract geometry Not robust Difficult to acquire illumination effects, e.g. specular highlights

More information

Introduction to object recognition. Slides adapted from Fei-Fei Li, Rob Fergus, Antonio Torralba, and others

Introduction to object recognition. Slides adapted from Fei-Fei Li, Rob Fergus, Antonio Torralba, and others Introduction to object recognition Slides adapted from Fei-Fei Li, Rob Fergus, Antonio Torralba, and others Overview Basic recognition tasks A statistical learning approach Traditional or shallow recognition

More information

Support Vector Machines

Support Vector Machines Support Vector Machines Michael Tagare De Guzman May 19, 2012 Support Vector Machines Linear Learning Machines and The Maximal Margin Classifier In Supervised Learning, a learning machine is given a training

More information

Gesture Recognition using Temporal Templates with disparity information

Gesture Recognition using Temporal Templates with disparity information 8- MVA7 IAPR Conference on Machine Vision Applications, May 6-8, 7, Tokyo, JAPAN Gesture Recognition using Temporal Templates with disparity information Kazunori Onoguchi and Masaaki Sato Hirosaki University

More information

Support Vector Machines.

Support Vector Machines. Support Vector Machines srihari@buffalo.edu SVM Discussion Overview 1. Overview of SVMs 2. Margin Geometry 3. SVM Optimization 4. Overlapping Distributions 5. Relationship to Logistic Regression 6. Dealing

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

Optic Flow and Basics Towards Horn-Schunck 1

Optic Flow and Basics Towards Horn-Schunck 1 Optic Flow and Basics Towards Horn-Schunck 1 Lecture 7 See Section 4.1 and Beginning of 4.2 in Reinhard Klette: Concise Computer Vision Springer-Verlag, London, 2014 1 See last slide for copyright information.

More information

Particle Filtering. CS6240 Multimedia Analysis. Leow Wee Kheng. Department of Computer Science School of Computing National University of Singapore

Particle Filtering. CS6240 Multimedia Analysis. Leow Wee Kheng. Department of Computer Science School of Computing National University of Singapore Particle Filtering CS6240 Multimedia Analysis Leow Wee Kheng Department of Computer Science School of Computing National University of Singapore (CS6240) Particle Filtering 1 / 28 Introduction Introduction

More information

Unsupervised Human Members Tracking Based on an Silhouette Detection and Analysis Scheme

Unsupervised Human Members Tracking Based on an Silhouette Detection and Analysis Scheme Unsupervised Human Members Tracking Based on an Silhouette Detection and Analysis Scheme Costas Panagiotakis and Anastasios Doulamis Abstract In this paper, an unsupervised, automatic video human members(human

More information

ECG782: Multidimensional Digital Signal Processing

ECG782: Multidimensional Digital Signal Processing ECG782: Multidimensional Digital Signal Processing Object Recognition http://www.ee.unlv.edu/~b1morris/ecg782/ 2 Outline Knowledge Representation Statistical Pattern Recognition Neural Networks Boosting

More information

Mouse Pointer Tracking with Eyes

Mouse Pointer Tracking with Eyes Mouse Pointer Tracking with Eyes H. Mhamdi, N. Hamrouni, A. Temimi, and M. Bouhlel Abstract In this article, we expose our research work in Human-machine Interaction. The research consists in manipulating

More information

Support Vector Machines

Support Vector Machines Support Vector Machines . Importance of SVM SVM is a discriminative method that brings together:. computational learning theory. previously known methods in linear discriminant functions 3. optimization

More information

Implementation of Optical Flow, Sliding Window and SVM for Vehicle Detection and Tracking

Implementation of Optical Flow, Sliding Window and SVM for Vehicle Detection and Tracking Implementation of Optical Flow, Sliding Window and SVM for Vehicle Detection and Tracking Mohammad Baji, Dr. I. SantiPrabha 2 M. Tech scholar, Department of E.C.E,U.C.E.K,Jawaharlal Nehru Technological

More information

I D I A P RECOGNITION OF ISOLATED COMPLEX MONO- AND BI-MANUAL 3D HAND GESTURES R E S E A R C H R E P O R T

I D I A P RECOGNITION OF ISOLATED COMPLEX MONO- AND BI-MANUAL 3D HAND GESTURES R E S E A R C H R E P O R T R E S E A R C H R E P O R T I D I A P RECOGNITION OF ISOLATED COMPLEX MONO- AND BI-MANUAL 3D HAND GESTURES Agnès Just a Olivier Bernier b Sébastien Marcel a IDIAP RR 03-63 FEBRUARY 2004 TO APPEAR IN Proceedings

More information

Visual Tracking (1) Feature Point Tracking and Block Matching

Visual Tracking (1) Feature Point Tracking and Block Matching Intelligent Control Systems Visual Tracking (1) Feature Point Tracking and Block Matching Shingo Kagami Graduate School of Information Sciences, Tohoku University swk(at)ic.is.tohoku.ac.jp http://www.ic.is.tohoku.ac.jp/ja/swk/

More information

RECOGNITION OF ISOLATED COMPLEX MONO- AND BI-MANUAL 3D HAND GESTURES USING DISCRETE IOHMM

RECOGNITION OF ISOLATED COMPLEX MONO- AND BI-MANUAL 3D HAND GESTURES USING DISCRETE IOHMM RECOGNITION OF ISOLATED COMPLEX MONO- AND BI-MANUAL 3D HAND GESTURES USING DISCRETE IOHMM Agnès Just, Sébastien Marcel Institut Dalle Molle d Intelligence Artificielle Perceptive (IDIAP) CH-1920 Martigny

More information

GENDER CLASSIFICATION USING SUPPORT VECTOR MACHINES

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

More information

Optimal Separating Hyperplane and the Support Vector Machine. Volker Tresp Summer 2018

Optimal Separating Hyperplane and the Support Vector Machine. Volker Tresp Summer 2018 Optimal Separating Hyperplane and the Support Vector Machine Volker Tresp Summer 2018 1 (Vapnik s) Optimal Separating Hyperplane Let s consider a linear classifier with y i { 1, 1} If classes are linearly

More information

Exploring Curve Fitting for Fingers in Egocentric Images

Exploring Curve Fitting for Fingers in Egocentric Images Exploring Curve Fitting for Fingers in Egocentric Images Akanksha Saran Robotics Institute, Carnegie Mellon University 16-811: Math Fundamentals for Robotics Final Project Report Email: asaran@andrew.cmu.edu

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

Local Image preprocessing (cont d)

Local Image preprocessing (cont d) Local Image preprocessing (cont d) 1 Outline - Edge detectors - Corner detectors - Reading: textbook 5.3.1-5.3.5 and 5.3.10 2 What are edges? Edges correspond to relevant features in the image. An edge

More information

CLASSIFICATION WITH RADIAL BASIS AND PROBABILISTIC NEURAL NETWORKS

CLASSIFICATION WITH RADIAL BASIS AND PROBABILISTIC NEURAL NETWORKS CLASSIFICATION WITH RADIAL BASIS AND PROBABILISTIC NEURAL NETWORKS CHAPTER 4 CLASSIFICATION WITH RADIAL BASIS AND PROBABILISTIC NEURAL NETWORKS 4.1 Introduction Optical character recognition is one of

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

Hand Gesture Recognition using Depth Data

Hand Gesture Recognition using Depth Data Hand Gesture Recognition using Depth Data Xia Liu Ohio State University Columbus OH 43210 Kikuo Fujimura Honda Research Institute USA Mountain View CA 94041 Abstract A method is presented for recognizing

More information

EE795: Computer Vision and Intelligent Systems

EE795: Computer Vision and Intelligent Systems EE795: Computer Vision and Intelligent Systems Spring 2012 TTh 17:30-18:45 WRI C225 Lecture 02 130124 http://www.ee.unlv.edu/~b1morris/ecg795/ 2 Outline Basics Image Formation Image Processing 3 Intelligent

More information

Mariya Zhariy. Uttendorf Introduction to Optical Flow. Mariya Zhariy. Introduction. Determining. Optical Flow. Results. Motivation Definition

Mariya Zhariy. Uttendorf Introduction to Optical Flow. Mariya Zhariy. Introduction. Determining. Optical Flow. Results. Motivation Definition to Constraint to Uttendorf 2005 Contents to Constraint 1 Contents to Constraint 1 2 Constraint Contents to Constraint 1 2 Constraint 3 Visual cranial reflex(vcr)(?) to Constraint Rapidly changing scene

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

FAST AND RELIABLE RECOGNITION OF HUMAN MOTION FROM MOTION TRAJECTORIES USING WAVELET ANALYSIS

FAST AND RELIABLE RECOGNITION OF HUMAN MOTION FROM MOTION TRAJECTORIES USING WAVELET ANALYSIS FAST AND RELIABLE RECOGNITION OF HUMAN MOTION FROM MOTION TRAJECTORIES USING WAVELET ANALYSIS Shu-Fai WONG 1 and Kwan-Yee Kenneth WONG 1 1 Department of Computer Science and Information Systems, The University

More information

Module 4. Non-linear machine learning econometrics: Support Vector Machine

Module 4. Non-linear machine learning econometrics: Support Vector Machine Module 4. Non-linear machine learning econometrics: Support Vector Machine THE CONTRACTOR IS ACTING UNDER A FRAMEWORK CONTRACT CONCLUDED WITH THE COMMISSION Introduction When the assumption of linearity

More information

MIME: A Gesture-Driven Computer Interface

MIME: A Gesture-Driven Computer Interface MIME: A Gesture-Driven Computer Interface Daniel Heckenberg a and Brian C. Lovell b a Department of Computer Science and Electrical Engineering, The University of Queensland, Brisbane, Australia, 4072

More information

Edge and local feature detection - 2. Importance of edge detection in computer vision

Edge and local feature detection - 2. Importance of edge detection in computer vision Edge and local feature detection Gradient based edge detection Edge detection by function fitting Second derivative edge detectors Edge linking and the construction of the chain graph Edge and local feature

More information

Real time facial expression recognition from image sequences using Support Vector Machines

Real time facial expression recognition from image sequences using Support Vector Machines Real time facial expression recognition from image sequences using Support Vector Machines I. Kotsia a and I. Pitas a a Aristotle University of Thessaloniki, Department of Informatics, Box 451, 54124 Thessaloniki,

More information

HAND-GESTURE BASED FILM RESTORATION

HAND-GESTURE BASED FILM RESTORATION HAND-GESTURE BASED FILM RESTORATION Attila Licsár University of Veszprém, Department of Image Processing and Neurocomputing,H-8200 Veszprém, Egyetem u. 0, Hungary Email: licsara@freemail.hu Tamás Szirányi

More information

Support Vector Machines + Classification for IR

Support Vector Machines + Classification for IR Support Vector Machines + Classification for IR Pierre Lison University of Oslo, Dep. of Informatics INF3800: Søketeknologi April 30, 2014 Outline of the lecture Recap of last week Support Vector Machines

More information

Classification by Nearest Shrunken Centroids and Support Vector Machines

Classification by Nearest Shrunken Centroids and Support Vector Machines Classification by Nearest Shrunken Centroids and Support Vector Machines Florian Markowetz florian.markowetz@molgen.mpg.de Max Planck Institute for Molecular Genetics, Computational Diagnostics Group,

More information

Robotics Programming Laboratory

Robotics Programming Laboratory Chair of Software Engineering Robotics Programming Laboratory Bertrand Meyer Jiwon Shin Lecture 8: Robot Perception Perception http://pascallin.ecs.soton.ac.uk/challenges/voc/databases.html#caltech car

More information

Support vector machines

Support vector machines Support vector machines When the data is linearly separable, which of the many possible solutions should we prefer? SVM criterion: maximize the margin, or distance between the hyperplane and the closest

More information

Wikipedia - Mysid

Wikipedia - Mysid Wikipedia - Mysid Erik Brynjolfsson, MIT Filtering Edges Corners Feature points Also called interest points, key points, etc. Often described as local features. Szeliski 4.1 Slides from Rick Szeliski,

More information

Classification by Support Vector Machines

Classification by Support Vector Machines Classification by Support Vector Machines Florian Markowetz Max-Planck-Institute for Molecular Genetics Computational Molecular Biology Berlin Practical DNA Microarray Analysis 2003 1 Overview I II III

More information

Robust Camera Pan and Zoom Change Detection Using Optical Flow

Robust Camera Pan and Zoom Change Detection Using Optical Flow Robust Camera and Change Detection Using Optical Flow Vishnu V. Makkapati Philips Research Asia - Bangalore Philips Innovation Campus, Philips Electronics India Ltd. Manyata Tech Park, Nagavara, Bangalore

More information

Visual Tracking (1) Tracking of Feature Points and Planar Rigid Objects

Visual Tracking (1) Tracking of Feature Points and Planar Rigid Objects Intelligent Control Systems Visual Tracking (1) Tracking of Feature Points and Planar Rigid Objects Shingo Kagami Graduate School of Information Sciences, Tohoku University swk(at)ic.is.tohoku.ac.jp http://www.ic.is.tohoku.ac.jp/ja/swk/

More information

Dense Image-based Motion Estimation Algorithms & Optical Flow

Dense Image-based Motion Estimation Algorithms & Optical Flow Dense mage-based Motion Estimation Algorithms & Optical Flow Video A video is a sequence of frames captured at different times The video data is a function of v time (t) v space (x,y) ntroduction to motion

More information

Corner Detection. Harvey Rhody Chester F. Carlson Center for Imaging Science Rochester Institute of Technology

Corner Detection. Harvey Rhody Chester F. Carlson Center for Imaging Science Rochester Institute of Technology Corner Detection Harvey Rhody Chester F. Carlson Center for Imaging Science Rochester Institute of Technology rhody@cis.rit.edu April 11, 2006 Abstract Corners and edges are two of the most important geometrical

More information

SVM in Analysis of Cross-Sectional Epidemiological Data Dmitriy Fradkin. April 4, 2005 Dmitriy Fradkin, Rutgers University Page 1

SVM in Analysis of Cross-Sectional Epidemiological Data Dmitriy Fradkin. April 4, 2005 Dmitriy Fradkin, Rutgers University Page 1 SVM in Analysis of Cross-Sectional Epidemiological Data Dmitriy Fradkin April 4, 2005 Dmitriy Fradkin, Rutgers University Page 1 Overview The goals of analyzing cross-sectional data Standard methods used

More information

Indian Institute of Technology Kanpur District : Kanpur Team number: 2. Prof. A. K. Chaturvedi SANKET

Indian Institute of Technology Kanpur District : Kanpur Team number: 2. Prof. A. K. Chaturvedi SANKET CSIDC 2003 Interim Report Country : India University : Indian Institute of Technology Kanpur District : Kanpur 208016 Team number: 2 Mentor: Prof. A. K. Chaturvedi akc@iitk.ac.in +91-512-2597613 SANKET:

More information

BioTechnology. An Indian Journal FULL PAPER. Trade Science Inc. Research on motion tracking and detection of computer vision ABSTRACT KEYWORDS

BioTechnology. An Indian Journal FULL PAPER. Trade Science Inc. Research on motion tracking and detection of computer vision ABSTRACT KEYWORDS [Type text] [Type text] [Type text] ISSN : 0974-7435 Volume 10 Issue 21 BioTechnology 2014 An Indian Journal FULL PAPER BTAIJ, 10(21), 2014 [12918-12922] Research on motion tracking and detection of computer

More information

Moving Object Segmentation Method Based on Motion Information Classification by X-means and Spatial Region Segmentation

Moving Object Segmentation Method Based on Motion Information Classification by X-means and Spatial Region Segmentation IJCSNS International Journal of Computer Science and Network Security, VOL.13 No.11, November 2013 1 Moving Object Segmentation Method Based on Motion Information Classification by X-means and Spatial

More information

An Approach for Reduction of Rain Streaks from a Single Image

An Approach for Reduction of Rain Streaks from a Single Image An Approach for Reduction of Rain Streaks from a Single Image Vijayakumar Majjagi 1, Netravati U M 2 1 4 th Semester, M. Tech, Digital Electronics, Department of Electronics and Communication G M Institute

More information

Traffic Signs Recognition using HP and HOG Descriptors Combined to MLP and SVM Classifiers

Traffic Signs Recognition using HP and HOG Descriptors Combined to MLP and SVM Classifiers Traffic Signs Recognition using HP and HOG Descriptors Combined to MLP and SVM Classifiers A. Salhi, B. Minaoui, M. Fakir, H. Chakib, H. Grimech Faculty of science and Technology Sultan Moulay Slimane

More information

Edge and corner detection

Edge and corner detection Edge and corner detection Prof. Stricker Doz. G. Bleser Computer Vision: Object and People Tracking Goals Where is the information in an image? How is an object characterized? How can I find measurements

More information

Lab 2: Support vector machines

Lab 2: Support vector machines Artificial neural networks, advanced course, 2D1433 Lab 2: Support vector machines Martin Rehn For the course given in 2006 All files referenced below may be found in the following directory: /info/annfk06/labs/lab2

More information

Motion in 2D image sequences

Motion in 2D image sequences Motion in 2D image sequences Definitely used in human vision Object detection and tracking Navigation and obstacle avoidance Analysis of actions or activities Segmentation and understanding of video sequences

More information

Face Recognition Using Vector Quantization Histogram and Support Vector Machine Classifier Rong-sheng LI, Fei-fei LEE *, Yan YAN and Qiu CHEN

Face Recognition Using Vector Quantization Histogram and Support Vector Machine Classifier Rong-sheng LI, Fei-fei LEE *, Yan YAN and Qiu CHEN 2016 International Conference on Artificial Intelligence: Techniques and Applications (AITA 2016) ISBN: 978-1-60595-389-2 Face Recognition Using Vector Quantization Histogram and Support Vector Machine

More information

Motion Estimation. There are three main types (or applications) of motion estimation:

Motion Estimation. There are three main types (or applications) of motion estimation: Members: D91922016 朱威達 R93922010 林聖凱 R93922044 謝俊瑋 Motion Estimation There are three main types (or applications) of motion estimation: Parametric motion (image alignment) The main idea of parametric motion

More information

10-701/15-781, Fall 2006, Final

10-701/15-781, Fall 2006, Final -7/-78, Fall 6, Final Dec, :pm-8:pm There are 9 questions in this exam ( pages including this cover sheet). If you need more room to work out your answer to a question, use the back of the page and clearly

More information

CS 565 Computer Vision. Nazar Khan PUCIT Lectures 15 and 16: Optic Flow

CS 565 Computer Vision. Nazar Khan PUCIT Lectures 15 and 16: Optic Flow CS 565 Computer Vision Nazar Khan PUCIT Lectures 15 and 16: Optic Flow Introduction Basic Problem given: image sequence f(x, y, z), where (x, y) specifies the location and z denotes time wanted: displacement

More information