WHO MISSED THE CLASS? - UNIFYING MULTI-FACE DETECTION, TRACKING AND RECOGNITION IN VIDEOS. Yunxiang Mao, Haohan Li, Zhaozheng Yin

Size: px
Start display at page:

Download "WHO MISSED THE CLASS? - UNIFYING MULTI-FACE DETECTION, TRACKING AND RECOGNITION IN VIDEOS. Yunxiang Mao, Haohan Li, Zhaozheng Yin"

Transcription

1 WHO MISSED THE CLASS? - UNIFYING MULTI-FACE DETECTION, TRACKING AND RECOGNITION IN VIDEOS Yunxiang Mao, Haohan Li, Zhaozheng Yin Department of Computer Science Missouri University of Science and Technology, USA ym8r8@mst.edu, hl87c@mst.edu, yinz@mst.edu ABSTRACT We investigate the problem of checking class attendance by detecting, tracking and recognizing multiple student faces in classroom videos taken by instructors. Instead of recognizing each individual face independently, first, we perform multiobject tracking to associate detected faces (including false positives) into face tracklets (each tracklet contains multiple instances of the same individual with variations in pose, illumination etc.) and then we cluster the face instances in each tracklet into a small number of clusters, achieving sparse face representation with less redundancy. Then, we formulate a unified optimization problem to (a) identify false positive face tracklets; (b) link broken face tracklets belonging to the same person due to long occlusion; and (c) recognize the group of faces simultaneously with spatial and temporal context constraints in the video. We test the proposed method on Honda/UCSD database and real classroom scenarios. The high recognition performance achieved by recognizing a group of multi-instance tracklets simultaneously demonstrates that multi-face recognition is more accurate than recognizing each individual face independently. Index Terms Face detection, face tracking, face recognition, multiple object tracking, multiple object recognition. 1. INTRODUCTION We are interested here in the class attendance checking by recognizing student faces in classroom images/videos taken by the instructor. Face recognition in images has been intensively researched in the past three decades [17, 21]. Usually, the workflow of image-based face recognition systems performs in the following steps: detect the human face in an image, compare the detected face with a reference dataset of N faces and classify who she/he is. This typical workflow builds face models/features for each individual person independently and recognizes each individual face. However, achieving high performance face recognition by using this one face image versus N references strategy is still a This work was supported by the NSF grant CNS and Intelligent Systems Center of Missouri S&T. challenging problem due to the variation of facial expression, pose, illumination condition, occlusion etc. In group photographs such as taking a short clip of video on students faces in a classroom, we can explore video information to improve the face recognition performance. In video-based face recognition, facial motion features are extracted for face recognition in addition to the appearancebased image features [6, 18]; 3D face models are fit into video frames to achieve pos-invariant face recognition [6, 10]; video dictionaries are designed to encode temporal and pose information for face recognition [2, 20]. Video-based face recognition is also converted into pairwise image set matching problem in [1, 4, 5, 16, 19]. Both the query and reference gallery contain a set of images for each individual corresponding to different poses, illuminations, expressions, etc. Faces from two videos are compared based on the similarity between two corresponding image sets. Considering the temporal recurrence of faces in a video volume, face tracking in videos has been adopted to improve face detection and face recognition. In [11], face tracking is leveraged to improve the hit rate of face detection in videos while decreasing the false alarm rate. A face tracker combining generative and discriminative face models [7] is developed to track faces in videos and build face models reflecting appearance variations, enabling face recognition by persondiscriminant features. By tracking faces gradually changing from the front view to profile view, the front and profile face appearances are linked and each identity contains more faces for recognition or age estimation [12, 13]. In [3], interdependencies between face tracking and face recognition is explored to improve the performance of multiple face recognition in videos. For the class attendance checking problem, we propose to improve the face recognition performance by using an Mversus-N strategy where M denotes the number of face sets detected in videos (note that the M sets may contain false positive face sets) and N denotes the number of reference faces (i.e., the number of registered students in the class). In the M-versus-N strategy, we recognize the M face sets simultaneously, considering the spatial and temporal context con-

2 Fig. 1. Overview of the proposed method. To protect the privacy of students, we hide their eyes in the paper. straints among them. The underlying motivation is that recognizing M face sets once is more accurate than recognizing individual face set M times. Fig.1 illustrates the workflow of the proposed system. Given a short clip of video taken on student faces in a classroom, face detector detects many faces candidates in individual images including non-face false positives. Then, we apply a multiple object tracking approach to link detected faces into tracklets based on the spatial-temporal information. Each tracklet contains multiple face instances with variations in pose, expression, illumination etc. The tracklets are summarized by sparse representations and encoded with entropyrelated features. Finally, we formulate a global optimization problem solving three problems at the same time: identify tracklets into reliable and unreliable tracklets such as false positives, connect broken tracklets which belong to the same person, and recognize the group of reliable tracklets. Our contributions in this paper include: (1) we perform multi-object tracking to associate detected faces (including false positives) into face tracklets (each tracklet contains multiple instances of the same individual with variations in pose, illumination etc.) and then we cluster the face instances in each face tracklet into a small number of clusters, achieving sparse face representation with less redundancy; (2) we formulate a unified optimization problem to: (a) identify false positive face tracklets; (b) link face tracklets belonging to the same person due to long occlusion; and (c) recognize the group of faces simultaneously with spatial and temporal context constraints in the video. 2. FACE DETECTION AND MULTI-FACE TRACKING 2.1. Face Detection In a crowded scene such as the classroom, some students may be occluded by other students who sit in front of them, thus it is hard to detect their faces from a single viewpoint. However, when the instructor moves from one end to the other end of the blackboard while recording videos, occluded faces will eventually appear. For each frame in the video, we apply the cascaded face detector based on Haar-like features [15] to detect faces, which works fine in the classroom scene (Fig.2). We largely decrease the miss detection rate to make sure that every person has her/his face detected in some images for the final recognition. The cost of low miss detection rate is to have quite some false positives. After this face detection step, we have thousands of cropped face images consisting of true positives and false positives. Fig. 2. Sample face detection results (detected faces are marked by red rectangles). The right column shows a student face before, during and after occlusion Face Tracking Recognizing each detected instance by comparing it with the reference gallery is not reliable due to the large variations in face pose, expression etc. In this paper, we group detected instances which belong to the same object into tracklets based on their temporal and spatial information. The face recognizer will recognize objects by tracklets instead of single image.

3 Three cases are possible in the face tracking (or tracklet association): linking, appear and disappear. For a single detected face image, we consider it as a special tracklet whose length is only one. Linking. Two tracklets T i and T j which belong to the same object may be separated due to miss detection. We define the cost to link the two tracklets as c(t i T j ) = l e i l s j + t e i t s j + s e i s s j l t s (1) where i, j [1, N], N is the number of tracklets; l e i is the tail location of tracklet T i and l s j is the head location of tracklet T j ; t e i is the time of the last instance of tracklet T i and t s j is the time of the first instance of tracklet T j; s e i s s j is the scale change between the last instance of tracklet T i and the first instance of tracklet T j. The linking cost between the two tracklets can only be considered when: the spatial distance between the tail location of T i and the head location of T j does not exceed the maximal translation distance ( l ); tracklet T j appears after tracklet T i, and the time difference between the tail of tracklet T i and the head of tracklet T j cannot exceed the maximal allowed time interval ( t ); the scale change ratio between tracklet T i and tracklet T j cannot exceed the maximal allowed ratio ( s ). Appear. A tracklet in the current frame reappears when it cannot be connected to any previous tracklet. The cost for a tracklet appearance is c( T j ) = N N N c(t i T j ) c(t i T j ) (2) i=1 j=1 i=1 arg min b c T b, s.t., R T b = 1 (4) where b is a binary vector in which b i = 1 means that the ith association hypothesis is selected in the optimal solution; R is a (NLinking T racklet + 2N T racklet ) 2N T racklet binary matrix, in which the nonzero elements of each row indicating which tracklets are involved in that association. By solving the binary integer programing in Eq.4, detected faces are grouped into tracklets. 3. RECOGNIZE A GROUP OF MULTI-INSTANCE TRACKLETS SIMULTANEOUSLY Recognizing a face tracklet will be more reliable than recognizing each individual face image since a tracklet contains multiple instances of an object with face variations. Instead of recognizing each tracklet independently, we can further improve the recognition performance by recognizing a group of tracklets simultaneously Recognizer Training To train a face recognizer, we perform our face detection and tracking on training videos and manually annotate the generated tracklets to construct a face image pool for each person. Within each face image pool, there are hundreds of faces with different illumination conditions, poses, and expressions. But since they are cropped in consecutive frames, many of them also share face similarities. We develop an effective way to reduce the redundancy in each image pool while still maintaining the face variations (Fig.3). First, we encode each image with the Histogram of Local Binary Pattern [9]. Then, we use K-mean to cluster face images in each pool (K=50 in our paper). Therefore, for each person, we have a small number of images for training. Finally, we train an eigen face recognizer [14] based on the collection of small sets of training images from all objects. Disappear. A tracklet in the previous frame disappears when it cannot be connected to any later tracklet either because it is occluded by some objects or it moves out of the field of view. The cost for a tracklet disappearance is c(t i ) = N N N c(t i T j ) c(t i T j ) (3) i=1 j=1 j=1 The number of linking hypothesis, denoted by NLinking T racklet, can be determined from the spatial and temporal constraints in Equation (1). Each tracklet can have disappearing and appearing hypotheses. Therefore, the dimension of a cost vector for all the possible associations is equal to NLinking T racklet + 2N T racklet. The following minimization problem can be formulated for tracklet association: Fig. 3. Generate sparse and representative training dataset.

4 3.2. Entropy-based Tracklet Feature arg max x ij M N+1 i=1 j=1 x ij e Dij (7) N+1 s.t., x ij 1, i [1, M] (8) j=1 (x pj + x qj )δ(c p Cq ) 1, j [1, N], p [1, M], q [1, M], p q (9) Fig. 4. Entropy encoding. Suppose we have N reference people in our training dataset, and each reference person has K reference images by K-mean clustering. Suppose after face tracking, we have M tracklets, and for the ith tracklet, there are m(i) detected face images, denoted by {q i1,..., q im(i) }. We project detected face image q ik (k [1, m(i)]) and reference image p jl (l [1, K]) into the eigen subspace and obtain their eigen coefficients. Then, we compute the norm-2 distance between their eigen coefficients, denoted by d ik,j l. We compute the distance between the kth instance of the ith tracklet and the jth reference person as d ik,j = min l d ik,j l, then we will have a 1 N vector, [d ik,1,, d ik,n ], indicating the distance between one tracklet instance and N reference people. To judge if one instance is informative or not, we compute its entropy as: f ik = N e d i k,j log(e d i k,j ) (5) j=1 where d i k,j is the normalized d i k,j by the maximum of d ik,j, j [1, K]. The entropy of a tracklet is computed as f i = m(i) k=1 f i k m(i) Fig.4 illustrates some entropy computation examples. If a tracklet is similar to one reference but dissimilar to other references, its entropy will be low (i.e., low uncertainty). Since false positives are not similar to every reference person, entropies of the instances in a false-positive tracklet should be high, resulting in high entropy in the tracklet. For some tracklets, if they are not similar to any reference people or they are similar to a set of reference people, these tracketlets will have high entropy indicating unreliable tracklets for recognition Global Tracklet Recognition We formulate the global tracklet recognition in a single optimization problem: (6) Objective function In Eq.7, x ij is a binary variable, D ij = min k,l d ik.j l. The (N+1)th reference represents unreliable trackets. The distance between the ith tracklet and the unreliable tracklets is defined as where E i = D i,n+1 = L i + E i + F i (10) { 0, m(i) < ML L i = 1, m(i) M L (11) { 1, minj (D ij ) < M E 0, min j (D ij ) M E (12) F i = { 1, fi < M F 0, f i M F (13) L i is a binary variable by thresholding the length of the ith tracklet. E i is a binary variable by thresholding the minimum distance between the ith tracklet and all reference people. F i is a binary variable by thresholding the entropy of the ith tracklet. Constraints The constraints in Eq.8 ensure that for each tracklet, it can only be assigned one label, either one of the reference people or unreliable tracklet. The constraints in Eq.9 consider the temporal exclusion among tracklets. A reference person s label cannot be assigned to two tracklets which have appeared together in at least one frame. Suppose the video has T frames, we construct an M T binary constraint matrix C, in which C ij = 1 means the ith tracklet appears in the jth frame. We scan from the first row to the last row of C. Each time we compare two rows (C p, C q ) to see if their intersection is empty. If C p Cq (i.e., δ(c p Cq ) = 1), labels from 1 to N cannot be assigned to tracklet p and tracklet q at the same time. By checking all pairwise rows, at most N(N 1) 2 constraints are added to the optimization problem. By solving the optimization function, tracklets are classified into two categorizes: reliable or unreliable tracklets; broken tracklets belonging to the same person due to long occlusion or large spatial distance/scale changes in videos can be

5 connected; the tracklets are recognized simultaneously compared to the reference images meanwhile the students who missed the class are also recognized since there is no tracklet associating to them. 4. EXPERIMENTS 4.1. Datasets The first dataset, Honda/UCSD, was collected by Lee et al. [8]. We take their first subset for our experiment, in which there are 59 videos on 20 different people including 20 videos for training and 39 videos for testing. The individuals were asked to demonstrate different expressions and head movements. The second dataset considers a small group recognition. It consists of three testing videos taken in a graduate class. 16 students are supposed to attend this class. 9, 9 and 13 students appear in the three videos, respectively. The third dataset considers a bigger undergraduate class which is supposed to have 39 students. There are 35, 37, 37, 34, 32 and 38 students in 6 videos, respectively Compare Face Recognition by Single-Image, Face Tracklet and Tracklet Group on the Honda/UCSD dataset Each video in the Honda/UCSD dataset contains a single person (the left of Fig.5). First, we run the face detector on the 20 training videos and then we apply the LBP+K-mean method on those detected faces to construct the sparse training dataset for each person. Then, we train our eigen recognizer for 20 people. After detecting and tracking faces in the 39 testing videos, the faces are grouped into 19 test videos corresponding to 19 different testing people. The first experiment on Honda dataset is to recognize each individual image in the test videos using the eigen recognizer. Table 1 summarizes the performance of face recognizer by single image on all test videos. The accuracy (ACC) is defined as the percentage of how many individual images out of the video are correctly recognized. On average, the accuracy is about 75% for all images in the testing videos, which indicates that recognition based on one single image is not reliable. Since people in the test videos are trying to act with different expression and head poses, it causes a big challenge for single-image-based recognition. The second experiment on Honda dataset is to recognize person by face tracklets. If we recognize one tracklet instead of one single image, this overcomes the difficulty caused by Fig. 5. Test datasets. different variations. Out of the 19 face tracklets corresponding to 19 different people, only one is wrongly recognized and we achieve recognition accuracy of 95%, as shown in the second column in Table 2 with k = 1). In the third experiment, we exhaustively evaluate all the possible combinations of selecting k (k [1, 19]) tracklets out of 19 tracklets to form a tracklet group and recognize the group simultaneously using our global recognition method. Table 2 shows that the average accuracy of using global tracklets recognition on different number of tracklets from test videos. The average accuracy increases while more tracklets are recognized together. 100% accuracy is achieved when the 19 tracklets are recognized simultaneously. By comparing the results from the three experiments, we can observe that recognizing tracklets is more accurate than recognizing single images, and recognizing a group of tracklets can achieve even higher accuracy than recognition based on individual tracklets Face Recognition in Classrooms We test our proposed method on two different video datasets captured by two instructors. In our implementation, the M L,M E and M F are set to be 14, 3000, and 19 respectively. For a dataset of V videos, when choosing 1 out of V as the testing video, we use the rest as training videos. Three evaluation metrics are used to evaluate the recognition performance: (1) precision on tracklets - the percentage of detected and tracked tracklets that are correctly recognized; (2) recall on students attending the class - the percentage of students attending the class whose associated tracklets are all correct; (3) recall on students missing the class - the percentage of students missing the class who do not have any tracklet associated. The results on the two classes are summarized in Table 3 and 4, respectively. On average, precision and recall above 90% are achieved in both classes. 5. CONCLUSION We investigated the problem of checking class attendance by detecting, tracking and recognizing multiple student faces in videos taken by instructors. We perform multi-object tracking to associate detected faces into face tracklets and then we cluster the face instances in each face tracklet into a small number of clusters, achieving sparse face representation with less redundance. Then, we formulate a unified optimiza-

6 Table 1. Recognition based on single image. Video No avg Acc (%) Table 2. Recognition based on tracklet group consisting of k tracklets. k Acc(%) tion problem to (a) identify false positive face tracklets; (b) link broken face tracklets belong to the same person due to long occlusion; and (c) recognize the group of faces simultaneously with spatial and temporal context constraints in the video. Experiments on Honda/UCSD database proved that recognizing a group of multi-instance tracklets is more accurate than recognizing a single tracklet or a single image. The high recognition performance achieved in classroom scenarios shows its promising feasibility and effectiveness for video-based class attendance checking. Table 3. Recognition performance in an undergraduate class with 39 students. Prec: precision on tracklets; Rec1: recall on attendance; Rec2: recall on missing class. Seq. Attn# Prec(%) Rec1 (%) Rec2(%) Avg Table 4. Recognition performance in a graduate class with 16 students. Prec: precision on tracklets; Rec1: recall on attendance; Rec2: recall on missing class. Seq. Attn# Prec(%) Rec1(%) Rec2(%) Avg REFERENCES [1] H. Cevikalp and B. Triggs, Face Recognition Based on Image Sets, in Proc. CVPR, Jun [2] Y. Chen, et al., Dictionary-Based Face Recognition from Video, in Proc. ECCV, Oct [3] A. Cohen and V. Paclovic, An Efficient IP Approach to Constrained Multiple Face Tracking and Recognition, in Proc. ICCV, Nov [4] Z. Cui, et al., Image Sets Alignment for Video-based Face Recognition, in Proc. CVPR, Jun [5] J. Harguessn, et al., Fusing Face Recognition from Multiple Cameras, in Proc. WACV, Dec [6] J. Harguess, et al., Full-Motion Recovery from Multiple Video Cameras Applied to Face Tracking and Recognition, in Proc. ICCV, Nov [7] M. Kim, et al., Face Tracking and Recognition with Visual Constraints in Real-World Videos, in Proc. CVPR, Jun [8] K. C. Lee, et al., Video-based face recognition using prbabilistic appearance manifolds, in Proc. CVPR, Jun [9] T. Ojala, et al., A Comparative Study of Texture Measures with Classification Based on Feature Distributions, Pattern Recognition, 29:51-59, [10] B. Raducanu and F. Dornaika, Pose-Invariant Face Recognition in Videos for Human-Machine Interaction, in Proc. ECCV, Oct [11] E. Sariyanidi, et al., LZM in Action: Realtime Face Recognition System, in Proc. ECCV, Oct [12] J. Sivic, et al., Who are you? - Learning person specific classifiers from video, in Proc. CVPR, Jun [13] Z. Song, et al., Learning universal multi-view age estimator by video contexts, in Proc. ICCV, NOV [14] M. A. Turk and A. P. Pentland, Face Recognition Using Eigenfaces, in Proc. CVPR, Jun [15] P. Viola and M. Jones, Rapid Object Detection using a Boosted Cascade of Simple Features, in Proc. CVPR, [16] R. Wang, et al., Manifold-Manifold Distance with application to face recognition based on image set, in Proc. CVPR, Jun [17] M. Yang, et al., Detecting Faces in Images: A Survey, in Proc. PAMI, [18] N. Ye and T. Sim, Towards general motion-based face recognition, in Proc. CVPR, [19] Q. Yin, et al., An Associate-Predict Model for Face Recognition, in Proc. CVPR, Jun [20] Q. Zhang and B. X. Li, Discriminative K-SVD for Dictionary Learning in Face Recognition, in Proc. CVPR, Jun [21] W. Zhao, et al., Face recognition: A literature survey, in ACM Computing Surveys, 2003.

Definition, Detection, and Evaluation of Meeting Events in Airport Surveillance Videos

Definition, Detection, and Evaluation of Meeting Events in Airport Surveillance Videos Definition, Detection, and Evaluation of Meeting Events in Airport Surveillance Videos Sung Chun Lee, Chang Huang, and Ram Nevatia University of Southern California, Los Angeles, CA 90089, USA sungchun@usc.edu,

More information

Object detection using non-redundant local Binary Patterns

Object detection using non-redundant local Binary Patterns University of Wollongong Research Online Faculty of Informatics - Papers (Archive) Faculty of Engineering and Information Sciences 2010 Object detection using non-redundant local Binary Patterns Duc Thanh

More information

Multiple Kernel Learning for Emotion Recognition in the Wild

Multiple Kernel Learning for Emotion Recognition in the Wild Multiple Kernel Learning for Emotion Recognition in the Wild Karan Sikka, Karmen Dykstra, Suchitra Sathyanarayana, Gwen Littlewort and Marian S. Bartlett Machine Perception Laboratory UCSD EmotiW Challenge,

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

FACE DETECTION AND RECOGNITION OF DRAWN CHARACTERS HERMAN CHAU

FACE DETECTION AND RECOGNITION OF DRAWN CHARACTERS HERMAN CHAU FACE DETECTION AND RECOGNITION OF DRAWN CHARACTERS HERMAN CHAU 1. Introduction Face detection of human beings has garnered a lot of interest and research in recent years. There are quite a few relatively

More information

Recognition of Animal Skin Texture Attributes in the Wild. Amey Dharwadker (aap2174) Kai Zhang (kz2213)

Recognition of Animal Skin Texture Attributes in the Wild. Amey Dharwadker (aap2174) Kai Zhang (kz2213) Recognition of Animal Skin Texture Attributes in the Wild Amey Dharwadker (aap2174) Kai Zhang (kz2213) Motivation Patterns and textures are have an important role in object description and understanding

More information

Partial Least Squares Regression on Grassmannian Manifold for Emotion Recognition

Partial Least Squares Regression on Grassmannian Manifold for Emotion Recognition Emotion Recognition In The Wild Challenge and Workshop (EmotiW 2013) Partial Least Squares Regression on Grassmannian Manifold for Emotion Recognition Mengyi Liu, Ruiping Wang, Zhiwu Huang, Shiguang Shan,

More information

Face Image Quality Assessment for Face Selection in Surveillance Video using Convolutional Neural Networks

Face Image Quality Assessment for Face Selection in Surveillance Video using Convolutional Neural Networks Face Image Quality Assessment for Face Selection in Surveillance Video using Convolutional Neural Networks Vignesh Sankar, K. V. S. N. L. Manasa Priya, Sumohana Channappayya Indian Institute of Technology

More information

2 Cascade detection and tracking

2 Cascade detection and tracking 3rd International Conference on Multimedia Technology(ICMT 213) A fast on-line boosting tracking algorithm based on cascade filter of multi-features HU Song, SUN Shui-Fa* 1, MA Xian-Bing, QIN Yin-Shi,

More information

Lecture 4 Face Detection and Classification. Lin ZHANG, PhD School of Software Engineering Tongji University Spring 2018

Lecture 4 Face Detection and Classification. Lin ZHANG, PhD School of Software Engineering Tongji University Spring 2018 Lecture 4 Face Detection and Classification Lin ZHANG, PhD School of Software Engineering Tongji University Spring 2018 Any faces contained in the image? Who are they? Outline Overview Face detection Introduction

More information

A NOVEL APPROACH TO ACCESS CONTROL BASED ON FACE RECOGNITION

A NOVEL APPROACH TO ACCESS CONTROL BASED ON FACE RECOGNITION A NOVEL APPROACH TO ACCESS CONTROL BASED ON FACE RECOGNITION A. Hadid, M. Heikkilä, T. Ahonen, and M. Pietikäinen Machine Vision Group Infotech Oulu and Department of Electrical and Information Engineering

More information

Face Detection and Alignment. Prof. Xin Yang HUST

Face Detection and Alignment. Prof. Xin Yang HUST Face Detection and Alignment Prof. Xin Yang HUST Many slides adapted from P. Viola Face detection Face detection Basic idea: slide a window across image and evaluate a face model at every location Challenges

More information

Large-Scale Traffic Sign Recognition based on Local Features and Color Segmentation

Large-Scale Traffic Sign Recognition based on Local Features and Color Segmentation Large-Scale Traffic Sign Recognition based on Local Features and Color Segmentation M. Blauth, E. Kraft, F. Hirschenberger, M. Böhm Fraunhofer Institute for Industrial Mathematics, Fraunhofer-Platz 1,

More information

PEOPLE IN SEATS COUNTING VIA SEAT DETECTION FOR MEETING SURVEILLANCE

PEOPLE IN SEATS COUNTING VIA SEAT DETECTION FOR MEETING SURVEILLANCE PEOPLE IN SEATS COUNTING VIA SEAT DETECTION FOR MEETING SURVEILLANCE Hongyu Liang, Jinchen Wu, and Kaiqi Huang National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Science

More information

Face Recognition via Sparse Representation

Face Recognition via Sparse Representation Face Recognition via Sparse Representation John Wright, Allen Y. Yang, Arvind, S. Shankar Sastry and Yi Ma IEEE Trans. PAMI, March 2008 Research About Face Face Detection Face Alignment Face Recognition

More information

Pairwise Threshold for Gaussian Mixture Classification and its Application on Human Tracking Enhancement

Pairwise Threshold for Gaussian Mixture Classification and its Application on Human Tracking Enhancement Pairwise Threshold for Gaussian Mixture Classification and its Application on Human Tracking Enhancement Daegeon Kim Sung Chun Lee Institute for Robotics and Intelligent Systems University of Southern

More information

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

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

People detection in complex scene using a cascade of Boosted classifiers based on Haar-like-features

People detection in complex scene using a cascade of Boosted classifiers based on Haar-like-features People detection in complex scene using a cascade of Boosted classifiers based on Haar-like-features M. Siala 1, N. Khlifa 1, F. Bremond 2, K. Hamrouni 1 1. Research Unit in Signal Processing, Image Processing

More information

An Adaptive Threshold LBP Algorithm for Face Recognition

An Adaptive Threshold LBP Algorithm for Face Recognition An Adaptive Threshold LBP Algorithm for Face Recognition Xiaoping Jiang 1, Chuyu Guo 1,*, Hua Zhang 1, and Chenghua Li 1 1 College of Electronics and Information Engineering, Hubei Key Laboratory of Intelligent

More information

A Real Time Facial Expression Classification System Using Local Binary Patterns

A Real Time Facial Expression Classification System Using Local Binary Patterns A Real Time Facial Expression Classification System Using Local Binary Patterns S L Happy, Anjith George, and Aurobinda Routray Department of Electrical Engineering, IIT Kharagpur, India Abstract Facial

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

Part-based and local feature models for generic object recognition

Part-based and local feature models for generic object recognition Part-based and local feature models for generic object recognition May 28 th, 2015 Yong Jae Lee UC Davis Announcements PS2 grades up on SmartSite PS2 stats: Mean: 80.15 Standard Dev: 22.77 Vote on piazza

More information

AUTOMATIC VIDEO INDEXING

AUTOMATIC VIDEO INDEXING AUTOMATIC VIDEO INDEXING Itxaso Bustos Maite Frutos TABLE OF CONTENTS Introduction Methods Key-frame extraction Automatic visual indexing Shot boundary detection Video OCR Index in motion Image processing

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

Haresh D. Chande #, Zankhana H. Shah *

Haresh D. Chande #, Zankhana H. Shah * Illumination Invariant Face Recognition System Haresh D. Chande #, Zankhana H. Shah * # Computer Engineering Department, Birla Vishvakarma Mahavidyalaya, Gujarat Technological University, India * Information

More information

Detection of a Single Hand Shape in the Foreground of Still Images

Detection of a Single Hand Shape in the Foreground of Still Images CS229 Project Final Report Detection of a Single Hand Shape in the Foreground of Still Images Toan Tran (dtoan@stanford.edu) 1. Introduction This paper is about an image detection system that can detect

More information

Human detection using local shape and nonredundant

Human detection using local shape and nonredundant University of Wollongong Research Online Faculty of Informatics - Papers (Archive) Faculty of Engineering and Information Sciences 2010 Human detection using local shape and nonredundant binary patterns

More information

Multiple-Person Tracking by Detection

Multiple-Person Tracking by Detection http://excel.fit.vutbr.cz Multiple-Person Tracking by Detection Jakub Vojvoda* Abstract Detection and tracking of multiple person is challenging problem mainly due to complexity of scene and large intra-class

More information

Previously. Part-based and local feature models for generic object recognition. Bag-of-words model 4/20/2011

Previously. Part-based and local feature models for generic object recognition. Bag-of-words model 4/20/2011 Previously Part-based and local feature models for generic object recognition Wed, April 20 UT-Austin Discriminative classifiers Boosting Nearest neighbors Support vector machines Useful for object recognition

More information

Face Alignment Under Various Poses and Expressions

Face Alignment Under Various Poses and Expressions Face Alignment Under Various Poses and Expressions Shengjun Xin and Haizhou Ai Computer Science and Technology Department, Tsinghua University, Beijing 100084, China ahz@mail.tsinghua.edu.cn Abstract.

More information

Applications Video Surveillance (On-line or off-line)

Applications Video Surveillance (On-line or off-line) Face Face Recognition: Dimensionality Reduction Biometrics CSE 190-a Lecture 12 CSE190a Fall 06 CSE190a Fall 06 Face Recognition Face is the most common biometric used by humans Applications range from

More information

Improving an Object Detector and Extracting Regions using Superpixels

Improving an Object Detector and Extracting Regions using Superpixels Improving an Object Detector and Extracting Regions using Superpixels Guang Shu, Afshin Dehghan, Mubarak Shah Computer Vision Lab, University of Central Florida {gshu, adehghan, shah}@eecs.ucf.edu Abstract

More information

Probabilistic Location Recognition using Reduced Feature Set

Probabilistic Location Recognition using Reduced Feature Set Probabilistic Location Recognition using Reduced Feature Set Fayin Li and Jana Košecá Department of Computer Science George Mason University, Fairfax, VA 3 Email: {fli,oseca}@cs.gmu.edu Abstract The localization

More information

IMPROVED FACE RECOGNITION USING ICP TECHNIQUES INCAMERA SURVEILLANCE SYSTEMS. Kirthiga, M.E-Communication system, PREC, Thanjavur

IMPROVED FACE RECOGNITION USING ICP TECHNIQUES INCAMERA SURVEILLANCE SYSTEMS. Kirthiga, M.E-Communication system, PREC, Thanjavur IMPROVED FACE RECOGNITION USING ICP TECHNIQUES INCAMERA SURVEILLANCE SYSTEMS Kirthiga, M.E-Communication system, PREC, Thanjavur R.Kannan,Assistant professor,prec Abstract: Face Recognition is important

More information

Rushes Video Segmentation Using Semantic Features

Rushes Video Segmentation Using Semantic Features Rushes Video Segmentation Using Semantic Features Athina Pappa, Vasileios Chasanis, and Antonis Ioannidis Department of Computer Science and Engineering, University of Ioannina, GR 45110, Ioannina, Greece

More information

Mobile Human Detection Systems based on Sliding Windows Approach-A Review

Mobile Human Detection Systems based on Sliding Windows Approach-A Review Mobile Human Detection Systems based on Sliding Windows Approach-A Review Seminar: Mobile Human detection systems Njieutcheu Tassi cedrique Rovile Department of Computer Engineering University of Heidelberg

More information

Tracking Using Online Feature Selection and a Local Generative Model

Tracking Using Online Feature Selection and a Local Generative Model Tracking Using Online Feature Selection and a Local Generative Model Thomas Woodley Bjorn Stenger Roberto Cipolla Dept. of Engineering University of Cambridge {tew32 cipolla}@eng.cam.ac.uk Computer Vision

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

Human Detection. A state-of-the-art survey. Mohammad Dorgham. University of Hamburg

Human Detection. A state-of-the-art survey. Mohammad Dorgham. University of Hamburg Human Detection A state-of-the-art survey Mohammad Dorgham University of Hamburg Presentation outline Motivation Applications Overview of approaches (categorized) Approaches details References Motivation

More information

Face/Flesh Detection and Face Recognition

Face/Flesh Detection and Face Recognition Face/Flesh Detection and Face Recognition Linda Shapiro EE/CSE 576 1 What s Coming 1. Review of Bakic flesh detector 2. Fleck and Forsyth flesh detector 3. Details of Rowley face detector 4. The Viola

More information

Bus Detection and recognition for visually impaired people

Bus Detection and recognition for visually impaired people Bus Detection and recognition for visually impaired people Hangrong Pan, Chucai Yi, and Yingli Tian The City College of New York The Graduate Center The City University of New York MAP4VIP Outline Motivation

More information

Visuelle Perzeption für Mensch- Maschine Schnittstellen

Visuelle Perzeption für Mensch- Maschine Schnittstellen Visuelle Perzeption für Mensch- Maschine Schnittstellen Vorlesung, WS 2009 Prof. Dr. Rainer Stiefelhagen Dr. Edgar Seemann Institut für Anthropomatik Universität Karlsruhe (TH) http://cvhci.ira.uka.de

More information

DescriptorEnsemble: An Unsupervised Approach to Image Matching and Alignment with Multiple Descriptors

DescriptorEnsemble: An Unsupervised Approach to Image Matching and Alignment with Multiple Descriptors DescriptorEnsemble: An Unsupervised Approach to Image Matching and Alignment with Multiple Descriptors 林彥宇副研究員 Yen-Yu Lin, Associate Research Fellow 中央研究院資訊科技創新研究中心 Research Center for IT Innovation, Academia

More information

Image Based Feature Extraction Technique For Multiple Face Detection and Recognition in Color Images

Image Based Feature Extraction Technique For Multiple Face Detection and Recognition in Color Images Image Based Feature Extraction Technique For Multiple Face Detection and Recognition in Color Images 1 Anusha Nandigam, 2 A.N. Lakshmipathi 1 Dept. of CSE, Sir C R Reddy College of Engineering, Eluru,

More information

Category-level localization

Category-level localization Category-level localization Cordelia Schmid Recognition Classification Object present/absent in an image Often presence of a significant amount of background clutter Localization / Detection Localize object

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

FACE RECOGNITION BASED ON LOCAL DERIVATIVE TETRA PATTERN

FACE RECOGNITION BASED ON LOCAL DERIVATIVE TETRA PATTERN ISSN: 976-92 (ONLINE) ICTACT JOURNAL ON IMAGE AND VIDEO PROCESSING, FEBRUARY 27, VOLUME: 7, ISSUE: 3 FACE RECOGNITION BASED ON LOCAL DERIVATIVE TETRA PATTERN A. Geetha, M. Mohamed Sathik 2 and Y. Jacob

More information

Object Detection by 3D Aspectlets and Occlusion Reasoning

Object Detection by 3D Aspectlets and Occlusion Reasoning Object Detection by 3D Aspectlets and Occlusion Reasoning Yu Xiang University of Michigan Silvio Savarese Stanford University In the 4th International IEEE Workshop on 3D Representation and Recognition

More information

FAST HUMAN DETECTION USING TEMPLATE MATCHING FOR GRADIENT IMAGES AND ASC DESCRIPTORS BASED ON SUBTRACTION STEREO

FAST HUMAN DETECTION USING TEMPLATE MATCHING FOR GRADIENT IMAGES AND ASC DESCRIPTORS BASED ON SUBTRACTION STEREO FAST HUMAN DETECTION USING TEMPLATE MATCHING FOR GRADIENT IMAGES AND ASC DESCRIPTORS BASED ON SUBTRACTION STEREO Makoto Arie, Masatoshi Shibata, Kenji Terabayashi, Alessandro Moro and Kazunori Umeda Course

More information

COSC160: Detection and Classification. Jeremy Bolton, PhD Assistant Teaching Professor

COSC160: Detection and Classification. Jeremy Bolton, PhD Assistant Teaching Professor COSC160: Detection and Classification Jeremy Bolton, PhD Assistant Teaching Professor Outline I. Problem I. Strategies II. Features for training III. Using spatial information? IV. Reducing dimensionality

More information

Postprint.

Postprint. http://www.diva-portal.org Postprint This is the accepted version of a paper presented at 14th International Conference of the Biometrics Special Interest Group, BIOSIG, Darmstadt, Germany, 9-11 September,

More information

Generic Face Alignment Using an Improved Active Shape Model

Generic Face Alignment Using an Improved Active Shape Model Generic Face Alignment Using an Improved Active Shape Model Liting Wang, Xiaoqing Ding, Chi Fang Electronic Engineering Department, Tsinghua University, Beijing, China {wanglt, dxq, fangchi} @ocrserv.ee.tsinghua.edu.cn

More information

Fuzzy based Multiple Dictionary Bag of Words for Image Classification

Fuzzy based Multiple Dictionary Bag of Words for Image Classification Available online at www.sciencedirect.com Procedia Engineering 38 (2012 ) 2196 2206 International Conference on Modeling Optimisation and Computing Fuzzy based Multiple Dictionary Bag of Words for Image

More information

CHAPTER 5 MOTION DETECTION AND ANALYSIS

CHAPTER 5 MOTION DETECTION AND ANALYSIS CHAPTER 5 MOTION DETECTION AND ANALYSIS 5.1. Introduction: Motion processing is gaining an intense attention from the researchers with the progress in motion studies and processing competence. A series

More information

Adaptive Dynamic Space Time Warping for Real Time Sign Language Recognition

Adaptive Dynamic Space Time Warping for Real Time Sign Language Recognition Adaptive Dynamic Space Time Warping for Real Time Sign Language Recognition Sergio Escalera, Alberto Escudero, Petia Radeva, and Jordi Vitrià Computer Vision Center, Campus UAB, Edifici O, 08193, Bellaterra,

More information

Capturing People in Surveillance Video

Capturing People in Surveillance Video Capturing People in Surveillance Video Rogerio Feris, Ying-Li Tian, and Arun Hampapur IBM T.J. Watson Research Center PO BOX 704, Yorktown Heights, NY 10598 {rsferis,yltian,arunh}@us.ibm.com Abstract This

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

Video shot segmentation using late fusion technique

Video shot segmentation using late fusion technique Video shot segmentation using late fusion technique by C. Krishna Mohan, N. Dhananjaya, B.Yegnanarayana in Proc. Seventh International Conference on Machine Learning and Applications, 2008, San Diego,

More information

Object Category Detection: Sliding Windows

Object Category Detection: Sliding Windows 04/10/12 Object Category Detection: Sliding Windows Computer Vision CS 543 / ECE 549 University of Illinois Derek Hoiem Today s class: Object Category Detection Overview of object category detection Statistical

More information

Det De e t cting abnormal event n s Jaechul Kim

Det De e t cting abnormal event n s Jaechul Kim Detecting abnormal events Jaechul Kim Purpose Introduce general methodologies used in abnormality detection Deal with technical details of selected papers Abnormal events Easy to verify, but hard to describe

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

Robust Human Detection Under Occlusion by Integrating Face and Person Detectors

Robust Human Detection Under Occlusion by Integrating Face and Person Detectors Robust Human Detection Under Occlusion by Integrating Face and Person Detectors William Robson Schwartz, Raghuraman Gopalan 2, Rama Chellappa 2, and Larry S. Davis University of Maryland, Department of

More information

Informative Census Transform for Ver Resolution Image Representation. Author(s)Jeong, Sungmoon; Lee, Hosun; Chong,

Informative Census Transform for Ver Resolution Image Representation. Author(s)Jeong, Sungmoon; Lee, Hosun; Chong, JAIST Reposi https://dspace.j Title Informative Census Transform for Ver Resolution Image Representation Author(s)Jeong, Sungmoon; Lee, Hosun; Chong, Citation IEEE International Symposium on Robo Interactive

More information

Action Recognition in Video by Sparse Representation on Covariance Manifolds of Silhouette Tunnels

Action Recognition in Video by Sparse Representation on Covariance Manifolds of Silhouette Tunnels Action Recognition in Video by Sparse Representation on Covariance Manifolds of Silhouette Tunnels Kai Guo, Prakash Ishwar, and Janusz Konrad Department of Electrical & Computer Engineering Motivation

More information

Binary Principal Component Analysis

Binary Principal Component Analysis 1 Binary Principal Component Analysis Feng Tang and Hai Tao University of California, Santa Cruz tang tao@soe.ucsc.edu Abstract Efficient and compact representation of images is a fundamental problem in

More information

Human Upper Body Pose Estimation in Static Images

Human Upper Body Pose Estimation in Static Images 1. Research Team Human Upper Body Pose Estimation in Static Images Project Leader: Graduate Students: Prof. Isaac Cohen, Computer Science Mun Wai Lee 2. Statement of Project Goals This goal of this project

More information

Subject-Oriented Image Classification based on Face Detection and Recognition

Subject-Oriented Image Classification based on Face Detection and Recognition 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

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

A Hybrid Face Detection System using combination of Appearance-based and Feature-based methods

A Hybrid Face Detection System using combination of Appearance-based and Feature-based methods IJCSNS International Journal of Computer Science and Network Security, VOL.9 No.5, May 2009 181 A Hybrid Face Detection System using combination of Appearance-based and Feature-based methods Zahra Sadri

More information

Sparse Models in Image Understanding And Computer Vision

Sparse Models in Image Understanding And Computer Vision Sparse Models in Image Understanding And Computer Vision Jayaraman J. Thiagarajan Arizona State University Collaborators Prof. Andreas Spanias Karthikeyan Natesan Ramamurthy Sparsity Sparsity of a vector

More information

arxiv: v1 [cs.cv] 16 Nov 2015

arxiv: v1 [cs.cv] 16 Nov 2015 Coarse-to-fine Face Alignment with Multi-Scale Local Patch Regression Zhiao Huang hza@megvii.com Erjin Zhou zej@megvii.com Zhimin Cao czm@megvii.com arxiv:1511.04901v1 [cs.cv] 16 Nov 2015 Abstract Facial

More information

on learned visual embedding patrick pérez Allegro Workshop Inria Rhônes-Alpes 22 July 2015

on learned visual embedding patrick pérez Allegro Workshop Inria Rhônes-Alpes 22 July 2015 on learned visual embedding patrick pérez Allegro Workshop Inria Rhônes-Alpes 22 July 2015 Vector visual representation Fixed-size image representation High-dim (100 100,000) Generic, unsupervised: BoW,

More information

Detecting Pedestrians Using Patterns of Motion and Appearance (Viola & Jones) - Aditya Pabbaraju

Detecting Pedestrians Using Patterns of Motion and Appearance (Viola & Jones) - Aditya Pabbaraju Detecting Pedestrians Using Patterns of Motion and Appearance (Viola & Jones) - Aditya Pabbaraju Background We are adept at classifying actions. Easily categorize even with noisy and small images Want

More information

Saliency based Person Re-Identification in Video using Colour Features

Saliency based Person Re-Identification in Video using Colour Features GRD Journals- Global Research and Development Journal for Engineering Volume 1 Issue 10 September 2016 ISSN: 2455-5703 Saliency based Person Re-Identification in Video using Colour Features Srujy Krishna

More information

Automatic Parameter Adaptation for Multi-Object Tracking

Automatic Parameter Adaptation for Multi-Object Tracking Automatic Parameter Adaptation for Multi-Object Tracking Duc Phu CHAU, Monique THONNAT, and François BREMOND {Duc-Phu.Chau, Monique.Thonnat, Francois.Bremond}@inria.fr STARS team, INRIA Sophia Antipolis,

More information

Announcements. Recognition I. Gradient Space (p,q) What is the reflectance map?

Announcements. Recognition I. Gradient Space (p,q) What is the reflectance map? Announcements I HW 3 due 12 noon, tomorrow. HW 4 to be posted soon recognition Lecture plan recognition for next two lectures, then video and motion. Introduction to Computer Vision CSE 152 Lecture 17

More information

Exploiting scene constraints to improve object detection algorithms for industrial applications

Exploiting scene constraints to improve object detection algorithms for industrial applications Exploiting scene constraints to improve object detection algorithms for industrial applications PhD Public Defense Steven Puttemans Promotor: Toon Goedemé 2 A general introduction Object detection? Help

More information

Face Recognition based Only on Eyes Information and Local Binary Pattern

Face Recognition based Only on Eyes Information and Local Binary Pattern Face Recognition based Only on Eyes Information and Local Binary Pattern Francisco Rosario-Verde, Joel Perez-Siles, Luis Aviles-Brito, Jesus Olivares-Mercado, Karina Toscano-Medina, and Hector Perez-Meana

More information

Virtual Training Samples and CRC based Test Sample Reconstruction and Face Recognition Experiments Wei HUANG and Li-ming MIAO

Virtual Training Samples and CRC based Test Sample Reconstruction and Face Recognition Experiments Wei HUANG and Li-ming MIAO 7 nd International Conference on Computational Modeling, Simulation and Applied Mathematics (CMSAM 7) ISBN: 978--6595-499-8 Virtual raining Samples and CRC based est Sample Reconstruction and Face Recognition

More information

Face and Nose Detection in Digital Images using Local Binary Patterns

Face and Nose Detection in Digital Images using Local Binary Patterns Face and Nose Detection in Digital Images using Local Binary Patterns Stanko Kružić Post-graduate student University of Split, Faculty of Electrical Engineering, Mechanical Engineering and Naval Architecture

More information

SLIDING WINDOW BASED MICRO-EXPRESSION SPOTTING: A BENCHMARK

SLIDING WINDOW BASED MICRO-EXPRESSION SPOTTING: A BENCHMARK SLIDING WINDOW BASED MICRO-EXPRESSION SPOTTING: A BENCHMARK Thuong-Khanh Tran, Xiaopeng Hong, and Guoying Zhao Center for Machine Vision and Signal Analysis, University of Oulu, Finland, fkhanh.tran, xiaopeng.hong,

More information

Human Detection and Tracking for Video Surveillance: A Cognitive Science Approach

Human Detection and Tracking for Video Surveillance: A Cognitive Science Approach Human Detection and Tracking for Video Surveillance: A Cognitive Science Approach Vandit Gajjar gajjar.vandit.381@ldce.ac.in Ayesha Gurnani gurnani.ayesha.52@ldce.ac.in Yash Khandhediya khandhediya.yash.364@ldce.ac.in

More information

Selection of Scale-Invariant Parts for Object Class Recognition

Selection of Scale-Invariant Parts for Object Class Recognition Selection of Scale-Invariant Parts for Object Class Recognition Gy. Dorkó and C. Schmid INRIA Rhône-Alpes, GRAVIR-CNRS 655, av. de l Europe, 3833 Montbonnot, France fdorko,schmidg@inrialpes.fr Abstract

More information

Skin and Face Detection

Skin and Face Detection Skin and Face Detection Linda Shapiro EE/CSE 576 1 What s Coming 1. Review of Bakic flesh detector 2. Fleck and Forsyth flesh detector 3. Details of Rowley face detector 4. Review of the basic AdaBoost

More information

String distance for automatic image classification

String distance for automatic image classification String distance for automatic image classification Nguyen Hong Thinh*, Le Vu Ha*, Barat Cecile** and Ducottet Christophe** *University of Engineering and Technology, Vietnam National University of HaNoi,

More information

Object and Class Recognition I:

Object and Class Recognition I: Object and Class Recognition I: Object Recognition Lectures 10 Sources ICCV 2005 short courses Li Fei-Fei (UIUC), Rob Fergus (Oxford-MIT), Antonio Torralba (MIT) http://people.csail.mit.edu/torralba/iccv2005

More information

Real Time Person Detection and Tracking by Mobile Robots using RGB-D Images

Real Time Person Detection and Tracking by Mobile Robots using RGB-D Images Real Time Person Detection and Tracking by Mobile Robots using RGB-D Images Duc My Vo, Lixing Jiang and Andreas Zell Abstract Detecting and tracking humans are key problems for human-robot interaction.

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

Face Cyclographs for Recognition

Face Cyclographs for Recognition Face Cyclographs for Recognition Guodong Guo Department of Computer Science North Carolina Central University E-mail: gdguo@nccu.edu Charles R. Dyer Computer Sciences Department University of Wisconsin-Madison

More information

Learning based face hallucination techniques: A survey

Learning based face hallucination techniques: A survey Vol. 3 (2014-15) pp. 37-45. : A survey Premitha Premnath K Department of Computer Science & Engineering Vidya Academy of Science & Technology Thrissur - 680501, Kerala, India (email: premithakpnath@gmail.com)

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

Object. Radiance. Viewpoint v

Object. Radiance. Viewpoint v Fisher Light-Fields for Face Recognition Across Pose and Illumination Ralph Gross, Iain Matthews, and Simon Baker The Robotics Institute, Carnegie Mellon University 5000 Forbes Avenue, Pittsburgh, PA 15213

More information

Video Google faces. Josef Sivic, Mark Everingham, Andrew Zisserman. Visual Geometry Group University of Oxford

Video Google faces. Josef Sivic, Mark Everingham, Andrew Zisserman. Visual Geometry Group University of Oxford Video Google faces Josef Sivic, Mark Everingham, Andrew Zisserman Visual Geometry Group University of Oxford The objective Retrieve all shots in a video, e.g. a feature length film, containing a particular

More information

Heat Kernel Based Local Binary Pattern for Face Representation

Heat Kernel Based Local Binary Pattern for Face Representation JOURNAL OF LATEX CLASS FILES 1 Heat Kernel Based Local Binary Pattern for Face Representation Xi Li, Weiming Hu, Zhongfei Zhang, Hanzi Wang Abstract Face classification has recently become a very hot research

More information

Searching Video Collections:Part I

Searching Video Collections:Part I Searching Video Collections:Part I Introduction to Multimedia Information Retrieval Multimedia Representation Visual Features (Still Images and Image Sequences) Color Texture Shape Edges Objects, Motion

More information

Image Processing Pipeline for Facial Expression Recognition under Variable Lighting

Image Processing Pipeline for Facial Expression Recognition under Variable Lighting Image Processing Pipeline for Facial Expression Recognition under Variable Lighting Ralph Ma, Amr Mohamed ralphma@stanford.edu, amr1@stanford.edu Abstract Much research has been done in the field of automated

More information

Image Set-based Face Recognition: A Local Multi-Keypoint Descriptor-based Approach

Image Set-based Face Recognition: A Local Multi-Keypoint Descriptor-based Approach 2013 IEEE Conference on Computer Vision and Pattern Recognition Workshops Image Set-based Face Recognition: A Local Multi-Keypoint Descriptor-based Approach Na Liu 1, Meng-Hui Lim 2, Pong C. Yuen 2, and

More information

Appearance Manifold of Facial Expression

Appearance Manifold of Facial Expression Appearance Manifold of Facial Expression Caifeng Shan, Shaogang Gong and Peter W. McOwan Department of Computer Science Queen Mary, University of London, London E1 4NS, UK {cfshan, sgg, pmco}@dcs.qmul.ac.uk

More information

Selecting Models from Videos for Appearance-Based Face Recognition

Selecting Models from Videos for Appearance-Based Face Recognition Selecting Models from Videos for Appearance-Based Face Recognition Abdenour Hadid and Matti Pietikäinen Machine Vision Group Infotech Oulu and Department of Electrical and Information Engineering P.O.

More information