Statistical Steganalyis of Images Using Open Source Software
|
|
- Maria Freeman
- 5 years ago
- Views:
Transcription
1 Statstcal Steganalys of Images Usng Open Source Software Bhargav Kapa, Stefan A. Robla Department of Computer Scence Montclar State Unversty Montclar, NJ Abstract In ths paper we present a novel steganalytc tool based on statstcal pattern recognton. The man am of our project was to desgn and mplement a system able to classfy the mages nto ones wth no hdden message and steganographc mages usng classc pattern classfcaton technques such as Bayesan classfcaton and decson trees. Experments are conducted on a large data set of mages to determne the classfcaton algorthm that performs better by comparng classfcaton success and error rates n each case. We have employed Weka, a data-mnng tool developed n java for ths purpose. We have also developed an applcaton usng Weka Java lbrary for loadng the data of the Images and classfy the mages nto normal mages and steganographc mages. Ths applcaton runs a GUI(Graphcal User Interface) that enables the user to choose the classfer and other optons requred for the classfcaton. Our results are algned wth current state of the art research and have the advantage of usng open source software. Keywords- mage processng, wavelets, steganography, steganalyss I. INTRODUCTION (HEADING 1) The word steganography s derved from the Greek words steganos and graphen, whch mean covered and wrtng. Steganography s often confused wth cryptography snce both are used to protect mportant nformaton. The dfference between the two s that steganography nvolves hdng nformaton n such a way that one cannot easly fnd that nformaton s hdden at all [1]. Hstorcal steganography nvolved technques such as dsappearng nk or mcrodots. Modern steganography nvolves hdng data n computer fles [2]. It s farly easy to hde a secret message n a graphc fle wthout notceably alterng the vsble appearance of that fle. In recent years, steganography has emerged as an ncreasngly actve research area, wth nformaton beng mperceptbly hdden n mages, vdeo, and audo among others. Wth the wde avalablty of dgtal mages, and the hgh degree of redundancy present n them despte compresson, there has been an ncreased nterest n usng dgtal mages as cover-objects for the purpose of steganography. Steganalyss,.e. the scence of dentfyng steganographc data s thus ncreasng n mportance. Whle many approaches to steganalsys have been proposed recently, often such approaches come offered as commercal or freeware packages and do not allow any reasonable accuracy testng. In ths paper, we descrbe the desgn and mplementaton of a system able to classfy non modfed and steganographc mages usng classc pattern recognton technques such as Bayesan classfcaton and decson trees. II. STEGANOGRAPHY AND STEGANALYSIS A generc descrpton of the steganographc process can be gven usng the followng formula : cover_medum + hdden_data + stego_key = stego_medum (1) where the cover_medum s the fle n whch the data s hdden, whch may also be encrypted usng the stego_key.the resultng fle s the stego_medum. The cover_medum (and, thus, the stego_medum) are typcally mage or audo fles. For example the cover mage shown n Fg. 1. was embedded wth the text of the Declaraton of Independence (over 6,600 characters) to obtan the mage on the rght. For ths we used the freely avalable Cryptobola Software that s able to embed encrypted data nto jpeg mages [3]. If presented wth the stego-mage, a human would not be able to realze that the mage ncludes addtonal nformaton. Even comparng sde by sde the stego and the cover medum would probably not help. There are many steganographc technques, ncludng: concealng messages wthn the lowest bts of mages or sound fles, concealng data wthn encrypted data, and embedded pctures n vdeo materal etc. The smplest approach to hdng data wthn an mage fle s called least sgnfcant bt (LSB) nserton. In ths method, we can take the bnary representaton of the hdden_data and overwrte the LSB of each byte wthn the cover mage. If we are usng 24-bt color, the amount of change wll be mnmal and ndscernble to the human eye [4]. Fgure 1. Cover mage on the left, stego-mage on the rght /10/$ IEEE
2 Though t s not possble to vsually dstngush between the cover mage and the steganographc mage, they can be dstngushed wth the help of certan steganalyss technques. Even f the secret content s not revealed, the exstence of t can be detected snce modfyng the cover medum results n the change of ts statstcal propertes and these changes cause dstortons n the resultng steganographc medum s statstcal propertes. The dstortons obtaned can then be used to analyze and detect steganographc mages [5]. In recent tme, many steganographc algorthms have also been desgned. Some of them are based on fle type and others, whch are more wdely used are based on embeddng method. Inserton-based technques hde data n sectons of a fle that are gnored by the processng applcaton and do not modfy those bts that determne the contents of a fle that are relevant to an end-user. In a substtuton-based algorthm, the most nsgnfcant bts of nformaton that determne the meanngful content of the orgnal fle are replaced wth new data n a way that causes the least amount of dstorton [6]. Wthn steganographc technques, a specal concern s gven to compressed nformaton such as Jpeg pctures. The Jpeg fle format s compact and does not sgnfcantly degrade the qualty of an mage so t s frequently used on the nternet. The Jpeg format uses a dscrete cosne transform (DCT) to dentfy 64 DCT coeffcents n successve 8x8 pxel blocks. Of these quantzed coeffcents, the least sgnfcant bts are used to embed data. Because modfcatons to these bts affect pxel frequency as opposed to spatal structure no obvous dstorton s present [7]. Most Jpeg Steganography methods replace the least sgnfcant bts (LSB) of the frequency coeffcents (DCT) skppng 0s and 1s after quantzaton (e.g. J-Steg, JP Hde&Seek, OutGuess) [8]. Other methods decrement the coeffcents absolute values when the LSBs do not match, except coeffcents wth value zero (e.g. F3, F4, F5) [2]. Steganalyss s not an easy task because of the dversty of natural mages and the wde varaton of data embeddng algorthms. However, an orgnal cover medum and ts stegoverson (wth hdden message nsde) always dffer from each other n some aspects snce the cover medum s modfed durng the data embeddng. Although all embeddng technques utlze redundances n the cover mage for the embeddng process, they dffer on ther approach, and the mage type they operate on. For example more recent technques such as F5, Outguess, and Perturbed quantzaton embeddng operate on Jpeg mages by modfyng the DCT coeffcents. Although changng the DCT coeffcents wll cause unnotceable vsual artfacts, they do cause detectable statstcal changes. These statstcal changes are used by steganalyss technques to detect any embedded messages. Essentally there are two approaches to the problem of steganalyss. One s to come up wth a steganalyss method specfc to a partcular steganographc algorthm. The other s developng unversal steganalyss technques whch are ndependent of the steganographc algorthm beng analyzed [7]. Each of the two approaches have ther own advantages and dsadvantages. A steganalyss technque specfc to an embeddng method would gve very good results when tested only on that embeddng method, and mght fal on all other steganographc algorthms. On the other hand, a steganalyss method whch s ndependent of the embeddng algorthm mght perform less accurately overall but stll provde acceptable results on new and unseen embeddng algorthms. Based on whether an mage contans hdden message, mages can be classfed nto two classes: the mage wth no hdden message and the correspondng steganogrphc mage (the very mage but wth message hdden n t). Steganalyss can thus be consdered as a pattern recognton process to decde whch class a test mage belongs to. The key ssue for steganalyss just lke for pattern recognton s feature extracton. The features should be senstve to the data hdng process. In other words, the features should be rather dfferent for the mage wthout hdden message and for the steganographc mage. The larger the dfference, the better the features are. The features should be as general as possble,.e., they are effectve to all dfferent types of mages and dfferent data hdng schemes [2]. III. PROPOSED METHOD A. Overall Approach The Methodology that s adopted for the project mplementaton has the sequence flow as shown n Fg. 2. A collecton of stego and plan mages s beng frst processed for extracton of features. The features are then reduced n number through an extracton process and the remnangn values are fed to a classfer. B. Feature Selecton Jpeg steganographc algorthms usually embed message bts nto randomly chosen DCT coeffcents. Hence the hstogram of the DCT coeffcents of a stego mage s dfferent from the hstogram of the DCT coeffcents of the cover mage, allowng the selecton of the coeffcents as features. Fg. 3 provdes an example of a cover and stego mage wth ther correspondng DCT coeffcent hstograms. The features were computed usng the tool ImageJ. ImageJ s an open source mage processng tool wrtten n Java. It s downloadable on any computer wth Java 1.4 or later vrtual machne [9]. In the current format, our applcaton requres manual processng of each mage through ImageJ. However, t can be easly envsoned that the DCT and the hstogram coeffcents can be computed automatcally wthout human nterventon. Data Set Feature Extracton Classfcaton Result analyss Fgure 2. Classfcaton of steganographc mages
3 C x) x C ) C ) x) = (3) As x) s constant, eq. 2 reduces to maxmzng P x C ) C ) (4) ( Whle ths approach s computatonally expensve for large n, to ease the burden class condtonal ndependence s assumed resultng n: Fgure 3. Cover mage on the left, stego-mage on the rght and the correspondng DCT hstograms below. C. Classfcaton Classfcaton of the mages s done usng Weka [10]. Weka stands for Wakato Envronment for Knowledge Analyss. Weka s a collecton of machne learnng algorthms for data mnng tasks. The algorthms can ether be appled drectly to a dataset or called from Java code. Weka contans tools for data pre-processng, classfcaton, regresson, clusterng, assocaton rules, and vsualzaton. It s also wellsuted for developng new machne learnng schemes. Weka s developed by the machne-learnng group of Computer Scence Department, Unversty of Wakato, New Zealand. To classfy we used the standard supervsed classfcaton approach. Here, the set of mages s splt n tranng and valdaton subsets. The tranng subset together wth nformaton on whether they are cover or stego mages s used to tran the classfer. The task of the supervsed classfer s to predct the label (stego or cover) for any mage after havng seen a number of tranng examples (.e. pars of nput and target output). To acheve ths, the classfer has to generalze from the presented data to new possble nputs [11]. Two dfferent classfcaton approaches were employed. The frst uses Naïve Bayesan classfcaton whle the second uses decson trees. The naïve Bayes classfer s based on the Bayesan theorem. It s partcularly suted when dmensonalty of the nputs s hgh. Parameter estmaton for nave Bayes models uses the method of maxmum lkelhood. In spte of over smplfed assumptons, t often performs better n many complex real-world stuatons. The man advantage s that t requres a small amount of tranng data to estmate the parameters. In a nutshell, the classfer works as follows: Gven a set D of n dmensonal vectors x (x 1, x 2, x 3,..., x n ), and m classes : C 1, C 2,..., C m the Naïve Bayes classfer predcts x belongs to class C m f: C x) > C j x) for all j between 1 and m (2) The above condtonal probablty can be expressed usng the Bayes theorem: P C ) = n k = 1 x k C ) (x (5) Overall, naïve Bayes classfcaton performs extremely well for a large number of problems. It s smplcty ensures wdespread use and the relatvely small number of exemplars needed for tranng s another supportng factors. Nevertheless, recent studes have shown that Bayesan classfcaton s outperformed by most modern classfers [12]. The second approach s based on decson trees. A decson tree s a predctve model that maps observatons about an tem wth conclusons about ts target value. The machne learnng technque for nducng a decson tree from data s called decson tree learnng [13]. More descrptve names for such tree models are classfcaton tree (dscrete outcome) or regresson tree (contnuous outcome). In these tree structures, leaves represent classfcatons and branches represent conjunctons of features that lead to those classfcatons. The machne learnng technque for nducng a decson tree from data s called decson tree learnng, or decson trees. Decson trees are herarchcal tree structures that can be used to classfy based on a seres of questons (or rules) about the attrbutes of the class. The attrbutes of the classes can be any type of varables from bnary, nomnal, ordnal, and quanttatve values, whle the classes must be qualtatve type (categorcal or bnary, or ordnal). In short, gven a data of attrbutes together wth ts classes, a decson tree produces a sequence of rules (or seres of questons) that can be used to recognze the class. There are several most popular decson tree algorthms such as ID3, C4.5 and CART (classfcaton and regresson trees). In general, the actual decson tree algorthms are recursve. In our case we used C4.5 developed by Ross Qunlan [14], as an extenson of an earler approach (ID3). The decson trees generated by C4.5 can be used for classfcaton, and for ths reason, C4.5 s often referred to as a statstcal classfer. C4.5 bulds decson trees from a set of tranng data, usng the concept of nformaton entropy. IV. APPLICATION Weka Java lbrary s leveraged to develop the applcaton. All the classes needed for the applcaton are mported from Weka Java lbrary to our applcaton. Java Code s wrtten n Eclpse IDE to load the data nto the applcaton, buld the classfer and evaluate the classfer. GUI s also developed n
4 eclpse IDE. Three sets of classes are needed for the applcaton: loadng data, buldng the classfer and evaluatng t. Our GUI based nterface allows us to load the features for a set of mages and tran the classfer usng ether naïve Bayes or decson trees. Next, the applcaton allows classfcaton of new mages usng ether of the classfers. V. RESULTS A data set of 1400 mages s created by frst takng 900 mages from the web. To create the setgo mages, we devlopped a java applcaton that generated random length text and embedded t usng the F5 algorthm [15], resultng n 500 stego mages. Experments are conducted by repeatng 10 tmes the followng: splt the data n 70%-30% rato for tran/ test data that s 70% of the data s used for tranng the classfer and the remanng 30% s used for testng. The results obtaned are n Tables 1 and 2. The average results are provded n Fg. 4. Decson tree classfer performed consstently well by correctly classfyng up to 70%-73% of the mages. From the results obtaned from Weka as well as our applcaton we can say that the performance of decson tree classfer s better than naïve Bayes classfer. To further understand the ablty of each approach to classfy correctly we looked at the confuson matrx. A confuson matrx contans nformaton about actual and predcted classfcatons done by the classfcaton system [16]. Performance of the classfer can be evaluated usng the data n the matrx. The entres n the confusonn matrx have the followng meanng n the context of our study: a s the fracton of correct predctons that an nstance s negatve, b s the fracton of ncorrect predctons that an nstance s postve, c s the fracton of ncorrect predctons that an nstance negatve, and d s the fracton of correct predctons that an nstance s postve (see Table 3). Table 4 presents the confuson matrces for both approaches. The data suggest that naïve Bayes s based towards detecton of stego mages whle decson trees s more accurate for cover mages. VI. CONCLUSION A system that can classfy the mages nto mages wth no hdden message and steganographc mages usng classc pattern recognton technques such as Bayesan classfcaton and decson trees was desgned, developed and mplemented successfully. Experments were conducted on a large data set of mages to determne the classfcaton algorthm thatt performs better by comparng classfcaton success and error rates n each case. We have employed Weka, a data mnng tool developed n java for ths purpose. We have also developed an applcaton usng Weka Java lbrary for loadng the data of the Images and classfy the mages nto normal mages and steganographc mages. Ths applcaton runs a GUI (Graphcal User Interface) that enables the user to choose the classfer and other optons requred for the classfcaton. TABLE I. Correctly classfed Instances % % % % % % % % % TABLE II. Correctly classfed Instances % % % % % % % % % % Real Cover Real Stego 80.00% 60.00% 40.00% 20.00% 0.00% NAÏVE BAYES B RESULTS DECISION TREEE CLASSIFIER RESULTS Fgure 4. Average classfcaton results. TABLE III. Actual Predct Cover Correct TABLE IV. Negatve Postve Naïve Bayes Predct Stego 13.34% 86.66% 20.73% 79.27% Incorrect CONFUSION MATRIX Predcted Negatve Postve a b c Incorrectly classfed nstances % % % % % % % % % % Incorrectly classfed nstances % % % % % % % % % % CONFUSION MATRIX Steganography algorthms contnue to evolve and add new devatons such as ncreased encrypton strength, vared message length or random postonng of the data. At each step, steganalyss technques focused on a sngle approach become obsolete snce they cannot detect the new d Bayes Decson Tree Decson Tree Predct Cover Predct Stego 87.27% 12.73% 63.41% 36.59%
5 steganographc methods. Technques lke the one we provde are consdered blnd,.e. ndependent of the steganographcal approach, and hold the very exctng potental to transcend the current fragle nature of modern steganalyss. They allow the detecton that a fle has a hdden message, even f t s hdden usng a new, prevously unseen steganography algorthm snce some or the other alteratons are made by these algorthms after embeddng the data. Whle not hghly accurate, blnd steganalyss remans an rreplaceable tool n the chase wth recent steganography approaches. REFERENCES [1] Y. Wang and P. Mouln, Optmzed feature extracton for learnngbased mage steganalyss, IEEE Transactons on Informaton Forenscs and Securty, vol. 2, 2007, pp [2] X.Y. Luo, D.S. Wang, P. Wang, and F.L. Lu, A revew on blnd detecton for mage steganography, Sgnal Processng, [3] CryptoBola, CryptoBola, [4] J. Zhang, Y. Hu, and Z. Yuan, Detecton of LSB Matchng Steganography usng the Envelope of Hstogram, Journal of Computers, vol. 4, 2009, p [5] C. Zhou, J. Feng, and Y. Yang, Blnd Steganalyss Based on Features n Fractonal Fourer Transform Doman. [6] M. Wess, Prncples of Steganography. [7] M. Kharraz, H.T. Sencar, and N. Memon, Image steganography: Concepts and practce, Lecture Note Seres, Insttute for Mathematcal Scences, Natonal Unversty of Sngapore, [8] G. Berg, I. Davdson, M.Y. Duan, and G. Paul, Searchng for hdden messages: Automatc detecton of steganography, Proceedngs of the 15th Innovatve Applcatons of AI Conference, 2003, pp [9] M.D. Abramoff, P.J. Magalhaes, and S.J. Ram, Image processng wth ImageJ, Bophotoncs Internatonal, vol. 11, 2004, pp [10] I.H. Wtten, E. Frank, L. Trgg, M. Hall, G. Holmes, and S.J. Cunnngham, Weka: Practcal machne learnng tools and technques wth Java mplementatons, ICONIP/ANZIIS/ANNES, Cteseer, 1999, pp [11] S.B. Kotsants, Supervsed machne learnng: A revew of classfcaton technques, EDITORIAL BOARDS, PUBLISHING COUNCIL, vol. 31, 2007, pp [12] R. Caruana and A. Nculescu-Mzl, An emprcal comparson of supervsed learnng algorthms, Proceedngs of the 23rd nternatonal conference on Machne learnng, ACM, 2006, p [13] R.O. Duda, P.E. Hart, and D.G. Stork, Pattern classfcaton, Cteseer, [14] J.R. Qunlan, C4. 5: programs for machne learnng, Morgan Kaufmann, [15] J. Frdrch, M. Goljan, and D. Hogea, Steganalyss of JPEG mages: Breakng the F5 algorthm, Lecture notes n computer scence, 2003, pp [16] F. Provost, T. Fawcett, and R. Kohav, The case aganst accuracy estmaton for comparng nducton algorthms, Proceedngs of the Ffteenth Internatonal Conference on Machne Learnng, Cteseer, 1998, pp
A Binarization Algorithm specialized on Document Images and Photos
A Bnarzaton Algorthm specalzed on Document mages and Photos Ergna Kavalleratou Dept. of nformaton and Communcaton Systems Engneerng Unversty of the Aegean kavalleratou@aegean.gr Abstract n ths paper, a
More informationFEATURE EXTRACTION. Dr. K.Vijayarekha. Associate Dean School of Electrical and Electronics Engineering SASTRA University, Thanjavur
FEATURE EXTRACTION Dr. K.Vjayarekha Assocate Dean School of Electrcal and Electroncs Engneerng SASTRA Unversty, Thanjavur613 41 Jont Intatve of IITs and IISc Funded by MHRD Page 1 of 8 Table of Contents
More informationHybrid Non-Blind Color Image Watermarking
Hybrd Non-Blnd Color Image Watermarkng Ms C.N.Sujatha 1, Dr. P. Satyanarayana 2 1 Assocate Professor, Dept. of ECE, SNIST, Yamnampet, Ghatkesar Hyderabad-501301, Telangana 2 Professor, Dept. of ECE, AITS,
More informationSupport Vector Machines
/9/207 MIST.6060 Busness Intellgence and Data Mnng What are Support Vector Machnes? Support Vector Machnes Support Vector Machnes (SVMs) are supervsed learnng technques that analyze data and recognze patterns.
More informationSkew Angle Estimation and Correction of Hand Written, Textual and Large areas of Non-Textual Document Images: A Novel Approach
Angle Estmaton and Correcton of Hand Wrtten, Textual and Large areas of Non-Textual Document Images: A Novel Approach D.R.Ramesh Babu Pyush M Kumat Mahesh D Dhannawat PES Insttute of Technology Research
More informationA Secured Method for Image Steganography Based On Pixel Values
A Secured Method for Image Steganography Based On Pxel Values Tarun Gulat #, Sanskrt Gupta * # Assocate Professor, Electroncs and Communcaton Engneerng Department, MMEC, M.M.U., Mullana, Ambala, Haryana,
More informationUser Authentication Based On Behavioral Mouse Dynamics Biometrics
User Authentcaton Based On Behavoral Mouse Dynamcs Bometrcs Chee-Hyung Yoon Danel Donghyun Km Department of Computer Scence Department of Computer Scence Stanford Unversty Stanford Unversty Stanford, CA
More informationThe Research of Support Vector Machine in Agricultural Data Classification
The Research of Support Vector Machne n Agrcultural Data Classfcaton Le Sh, Qguo Duan, Xnmng Ma, Me Weng College of Informaton and Management Scence, HeNan Agrcultural Unversty, Zhengzhou 45000 Chna Zhengzhou
More informationCluster Analysis of Electrical Behavior
Journal of Computer and Communcatons, 205, 3, 88-93 Publshed Onlne May 205 n ScRes. http://www.scrp.org/ournal/cc http://dx.do.org/0.4236/cc.205.350 Cluster Analyss of Electrcal Behavor Ln Lu Ln Lu, School
More informationHigh Payload Reversible Data Hiding Scheme Using Difference Segmentation and Histogram Shifting
JOURNAL OF ELECTRONIC SCIENCE AND TECHNOLOGY, VOL. 11, NO. 1, MARCH 2013 9 Hgh Payload Reversble Data Hdng Scheme Usng Dfference Segmentaton and Hstogram Shftng Yung-Chen Chou and Huang-Chng L Abstract
More informationContent Based Image Retrieval Using 2-D Discrete Wavelet with Texture Feature with Different Classifiers
IOSR Journal of Electroncs and Communcaton Engneerng (IOSR-JECE) e-issn: 78-834,p- ISSN: 78-8735.Volume 9, Issue, Ver. IV (Mar - Apr. 04), PP 0-07 Content Based Image Retreval Usng -D Dscrete Wavelet wth
More informationS1 Note. Basis functions.
S1 Note. Bass functons. Contents Types of bass functons...1 The Fourer bass...2 B-splne bass...3 Power and type I error rates wth dfferent numbers of bass functons...4 Table S1. Smulaton results of type
More informationSLAM Summer School 2006 Practical 2: SLAM using Monocular Vision
SLAM Summer School 2006 Practcal 2: SLAM usng Monocular Vson Javer Cvera, Unversty of Zaragoza Andrew J. Davson, Imperal College London J.M.M Montel, Unversty of Zaragoza. josemar@unzar.es, jcvera@unzar.es,
More informationFeature Reduction and Selection
Feature Reducton and Selecton Dr. Shuang LIANG School of Software Engneerng TongJ Unversty Fall, 2012 Today s Topcs Introducton Problems of Dmensonalty Feature Reducton Statstc methods Prncpal Components
More informationA PATTERN RECOGNITION APPROACH TO IMAGE SEGMENTATION
1 THE PUBLISHING HOUSE PROCEEDINGS OF THE ROMANIAN ACADEMY, Seres A, OF THE ROMANIAN ACADEMY Volume 4, Number 2/2003, pp.000-000 A PATTERN RECOGNITION APPROACH TO IMAGE SEGMENTATION Tudor BARBU Insttute
More informationA Fast Content-Based Multimedia Retrieval Technique Using Compressed Data
A Fast Content-Based Multmeda Retreval Technque Usng Compressed Data Borko Furht and Pornvt Saksobhavvat NSF Multmeda Laboratory Florda Atlantc Unversty, Boca Raton, Florda 3343 ABSTRACT In ths paper,
More informationMachine Learning: Algorithms and Applications
14/05/1 Machne Learnng: Algorthms and Applcatons Florano Zn Free Unversty of Bozen-Bolzano Faculty of Computer Scence Academc Year 011-01 Lecture 10: 14 May 01 Unsupervsed Learnng cont Sldes courtesy of
More informationLearning the Kernel Parameters in Kernel Minimum Distance Classifier
Learnng the Kernel Parameters n Kernel Mnmum Dstance Classfer Daoqang Zhang 1,, Songcan Chen and Zh-Hua Zhou 1* 1 Natonal Laboratory for Novel Software Technology Nanjng Unversty, Nanjng 193, Chna Department
More informationEdge Detection in Noisy Images Using the Support Vector Machines
Edge Detecton n Nosy Images Usng the Support Vector Machnes Hlaro Gómez-Moreno, Saturnno Maldonado-Bascón, Francsco López-Ferreras Sgnal Theory and Communcatons Department. Unversty of Alcalá Crta. Madrd-Barcelona
More informationMULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION
MULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION Paulo Quntlano 1 & Antono Santa-Rosa 1 Federal Polce Department, Brasla, Brazl. E-mals: quntlano.pqs@dpf.gov.br and
More informationA Fast Visual Tracking Algorithm Based on Circle Pixels Matching
A Fast Vsual Trackng Algorthm Based on Crcle Pxels Matchng Zhqang Hou hou_zhq@sohu.com Chongzhao Han czhan@mal.xjtu.edu.cn Ln Zheng Abstract: A fast vsual trackng algorthm based on crcle pxels matchng
More informationClassifying Acoustic Transient Signals Using Artificial Intelligence
Classfyng Acoustc Transent Sgnals Usng Artfcal Intellgence Steve Sutton, Unversty of North Carolna At Wlmngton (suttons@charter.net) Greg Huff, Unversty of North Carolna At Wlmngton (jgh7476@uncwl.edu)
More informationImplementation Naïve Bayes Algorithm for Student Classification Based on Graduation Status
Internatonal Journal of Appled Busness and Informaton Systems ISSN: 2597-8993 Vol 1, No 2, September 2017, pp. 6-12 6 Implementaton Naïve Bayes Algorthm for Student Classfcaton Based on Graduaton Status
More informationLecture 5: Multilayer Perceptrons
Lecture 5: Multlayer Perceptrons Roger Grosse 1 Introducton So far, we ve only talked about lnear models: lnear regresson and lnear bnary classfers. We noted that there are functons that can t be represented
More informationKey-Selective Patchwork Method for Audio Watermarking
Internatonal Journal of Dgtal Content Technology and ts Applcatons Volume 4, Number 4, July 2010 Key-Selectve Patchwork Method for Audo Watermarkng 1 Ch-Man Pun, 2 Jng-Jng Jang 1, Frst and Correspondng
More informationDetection of an Object by using Principal Component Analysis
Detecton of an Object by usng Prncpal Component Analyss 1. G. Nagaven, 2. Dr. T. Sreenvasulu Reddy 1. M.Tech, Department of EEE, SVUCE, Trupath, Inda. 2. Assoc. Professor, Department of ECE, SVUCE, Trupath,
More informationInvestigating the Performance of Naïve- Bayes Classifiers and K- Nearest Neighbor Classifiers
Journal of Convergence Informaton Technology Volume 5, Number 2, Aprl 2010 Investgatng the Performance of Naïve- Bayes Classfers and K- Nearest Neghbor Classfers Mohammed J. Islam *, Q. M. Jonathan Wu,
More informationA Modified Median Filter for the Removal of Impulse Noise Based on the Support Vector Machines
A Modfed Medan Flter for the Removal of Impulse Nose Based on the Support Vector Machnes H. GOMEZ-MORENO, S. MALDONADO-BASCON, F. LOPEZ-FERRERAS, M. UTRILLA- MANSO AND P. GIL-JIMENEZ Departamento de Teoría
More informationSimulation: Solving Dynamic Models ABE 5646 Week 11 Chapter 2, Spring 2010
Smulaton: Solvng Dynamc Models ABE 5646 Week Chapter 2, Sprng 200 Week Descrpton Readng Materal Mar 5- Mar 9 Evaluatng [Crop] Models Comparng a model wth data - Graphcal, errors - Measures of agreement
More informationFor instance, ; the five basic number-sets are increasingly more n A B & B A A = B (1)
Secton 1.2 Subsets and the Boolean operatons on sets If every element of the set A s an element of the set B, we say that A s a subset of B, or that A s contaned n B, or that B contans A, and we wrte A
More informationTerm Weighting Classification System Using the Chi-square Statistic for the Classification Subtask at NTCIR-6 Patent Retrieval Task
Proceedngs of NTCIR-6 Workshop Meetng, May 15-18, 2007, Tokyo, Japan Term Weghtng Classfcaton System Usng the Ch-square Statstc for the Classfcaton Subtask at NTCIR-6 Patent Retreval Task Kotaro Hashmoto
More informationReducing Frame Rate for Object Tracking
Reducng Frame Rate for Object Trackng Pavel Korshunov 1 and We Tsang Oo 2 1 Natonal Unversty of Sngapore, Sngapore 11977, pavelkor@comp.nus.edu.sg 2 Natonal Unversty of Sngapore, Sngapore 11977, oowt@comp.nus.edu.sg
More informationClassifier Selection Based on Data Complexity Measures *
Classfer Selecton Based on Data Complexty Measures * Edth Hernández-Reyes, J.A. Carrasco-Ochoa, and J.Fco. Martínez-Trndad Natonal Insttute for Astrophyscs, Optcs and Electroncs, Lus Enrque Erro No.1 Sta.
More informationCHAPTER 3 SEQUENTIAL MINIMAL OPTIMIZATION TRAINED SUPPORT VECTOR CLASSIFIER FOR CANCER PREDICTION
48 CHAPTER 3 SEQUENTIAL MINIMAL OPTIMIZATION TRAINED SUPPORT VECTOR CLASSIFIER FOR CANCER PREDICTION 3.1 INTRODUCTION The raw mcroarray data s bascally an mage wth dfferent colors ndcatng hybrdzaton (Xue
More informationFace Recognition Based on SVM and 2DPCA
Vol. 4, o. 3, September, 2011 Face Recognton Based on SVM and 2DPCA Tha Hoang Le, Len Bu Faculty of Informaton Technology, HCMC Unversty of Scence Faculty of Informaton Scences and Engneerng, Unversty
More informationPerformance Analysis of Data Hiding in MPEG-4 AAC Audio *
TSINGHUA SCIENCE AND TECHNOLOGY ISSNll1007-0214ll07/21llpp55-61 Volume 14, Number 1, February 2009 Performance Analyss of Data Hdng n MPEG-4 AAC Audo * XU Shuzheng ( ) **, ZHANG Peng ( ), WANG Pengjun
More informationArticle Reversible Dual-Image-Based Hiding Scheme Using Block Folding Technique
Artcle Reversble Dual-Image-Based Hdng Scheme Usng Block Foldng Technque Tzu-Chuen Lu, * and Hu-Shh Leng Department of Informaton Management, Chaoyang Unversty of Technology, Tachung 4349, Tawan Department
More informationWishing you all a Total Quality New Year!
Total Qualty Management and Sx Sgma Post Graduate Program 214-15 Sesson 4 Vnay Kumar Kalakband Assstant Professor Operatons & Systems Area 1 Wshng you all a Total Qualty New Year! Hope you acheve Sx sgma
More informationData Mining: Model Evaluation
Data Mnng: Model Evaluaton Aprl 16, 2013 1 Issues: Evaluatng Classfcaton Methods Accurac classfer accurac: predctng class label predctor accurac: guessng value of predcted attrbutes Speed tme to construct
More informationEYE CENTER LOCALIZATION ON A FACIAL IMAGE BASED ON MULTI-BLOCK LOCAL BINARY PATTERNS
P.G. Demdov Yaroslavl State Unversty Anatoly Ntn, Vladmr Khryashchev, Olga Stepanova, Igor Kostern EYE CENTER LOCALIZATION ON A FACIAL IMAGE BASED ON MULTI-BLOCK LOCAL BINARY PATTERNS Yaroslavl, 2015 Eye
More informationSubspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points;
Subspace clusterng Clusterng Fundamental to all clusterng technques s the choce of dstance measure between data ponts; D q ( ) ( ) 2 x x = x x, j k = 1 k jk Squared Eucldean dstance Assumpton: All features
More informationBackpropagation: In Search of Performance Parameters
Bacpropagaton: In Search of Performance Parameters ANIL KUMAR ENUMULAPALLY, LINGGUO BU, and KHOSROW KAIKHAH, Ph.D. Computer Scence Department Texas State Unversty-San Marcos San Marcos, TX-78666 USA ae049@txstate.edu,
More informationTsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance
Tsnghua Unversty at TAC 2009: Summarzng Mult-documents by Informaton Dstance Chong Long, Mnle Huang, Xaoyan Zhu State Key Laboratory of Intellgent Technology and Systems, Tsnghua Natonal Laboratory for
More informationIdentify the Attack in Embedded Image with Steganalysis Detection Method by PSNR and RGB Intensity
Internatonal Journal of Computer Systems (ISSN: 394-1065), Volume 03 Issue 07, July, 016 Avalable at http://www.jcsonlne.com/ Identfy the Attack n Embedded Image wth Steganalyss Detecton Method by PSNR
More informationCS 534: Computer Vision Model Fitting
CS 534: Computer Vson Model Fttng Sprng 004 Ahmed Elgammal Dept of Computer Scence CS 534 Model Fttng - 1 Outlnes Model fttng s mportant Least-squares fttng Maxmum lkelhood estmaton MAP estmaton Robust
More informationAudio Content Classification Method Research Based on Two-step Strategy
(IJACSA) Internatonal Journal of Advanced Computer Scence and Applcatons, Audo Content Classfcaton Method Research Based on Two-step Strategy Sume Lang Department of Computer Scence and Technology Chongqng
More informationCompiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz
Compler Desgn Sprng 2014 Regster Allocaton Sample Exercses and Solutons Prof. Pedro C. Dnz USC / Informaton Scences Insttute 4676 Admralty Way, Sute 1001 Marna del Rey, Calforna 90292 pedro@s.edu Regster
More informationBOOSTING CLASSIFICATION ACCURACY WITH SAMPLES CHOSEN FROM A VALIDATION SET
1 BOOSTING CLASSIFICATION ACCURACY WITH SAMPLES CHOSEN FROM A VALIDATION SET TZU-CHENG CHUANG School of Electrcal and Computer Engneerng, Purdue Unversty, West Lafayette, Indana 47907 SAUL B. GELFAND School
More informationOutline. Discriminative classifiers for image recognition. Where in the World? A nearest neighbor recognition example 4/14/2011. CS 376 Lecture 22 1
4/14/011 Outlne Dscrmnatve classfers for mage recognton Wednesday, Aprl 13 Krsten Grauman UT-Austn Last tme: wndow-based generc obect detecton basc ppelne face detecton wth boostng as case study Today:
More informationPRÉSENTATIONS DE PROJETS
PRÉSENTATIONS DE PROJETS Rex Onlne (V. Atanasu) What s Rex? Rex s an onlne browser for collectons of wrtten documents [1]. Asde ths core functon t has however many other applcatons that make t nterestng
More informationClassification Based Mode Decisions for Video over Networks
Classfcaton Based Mode Decsons for Vdeo over Networks Deepak S. Turaga and Tsuhan Chen Advanced Multmeda Processng Lab Tranng data for Inter-Intra Decson Inter-Intra Decson Regons pdf 6 5 6 5 Energy 4
More informationAssignment # 2. Farrukh Jabeen Algorithms 510 Assignment #2 Due Date: June 15, 2009.
Farrukh Jabeen Algorthms 51 Assgnment #2 Due Date: June 15, 29. Assgnment # 2 Chapter 3 Dscrete Fourer Transforms Implement the FFT for the DFT. Descrbed n sectons 3.1 and 3.2. Delverables: 1. Concse descrpton
More informationRelated-Mode Attacks on CTR Encryption Mode
Internatonal Journal of Network Securty, Vol.4, No.3, PP.282 287, May 2007 282 Related-Mode Attacks on CTR Encrypton Mode Dayn Wang, Dongda Ln, and Wenlng Wu (Correspondng author: Dayn Wang) Key Laboratory
More informationThe Codesign Challenge
ECE 4530 Codesgn Challenge Fall 2007 Hardware/Software Codesgn The Codesgn Challenge Objectves In the codesgn challenge, your task s to accelerate a gven software reference mplementaton as fast as possble.
More informationA Unified Framework for Semantics and Feature Based Relevance Feedback in Image Retrieval Systems
A Unfed Framework for Semantcs and Feature Based Relevance Feedback n Image Retreval Systems Ye Lu *, Chunhu Hu 2, Xngquan Zhu 3*, HongJang Zhang 2, Qang Yang * School of Computng Scence Smon Fraser Unversty
More informationA Robust Webpage Information Hiding Method Based on the Slash of Tag
Advanced Engneerng Forum Onlne: 2012-09-26 ISSN: 2234-991X, Vols. 6-7, pp 361-366 do:10.4028/www.scentfc.net/aef.6-7.361 2012 Trans Tech Publcatons, Swtzerland A Robust Webpage Informaton Hdng Method Based
More informationA User Selection Method in Advertising System
Int. J. Communcatons, etwork and System Scences, 2010, 3, 54-58 do:10.4236/jcns.2010.31007 Publshed Onlne January 2010 (http://www.scrp.org/journal/jcns/). A User Selecton Method n Advertsng System Shy
More informationDetecting MP3Stego using Calibrated Side Information Features
2628 JOURNAL OF SOFTWARE, VOL. 8, NO. 10, OCTOBER 2013 Detectng P3Stego usng Calbrated Sde Informaton Features Xanmn Yu School of Informaton Scence and Engneerng, Nngbo Unversty Emal: mlhappy1016@163.com
More informationAn Entropy-Based Approach to Integrated Information Needs Assessment
Dstrbuton Statement A: Approved for publc release; dstrbuton s unlmted. An Entropy-Based Approach to ntegrated nformaton Needs Assessment June 8, 2004 Wllam J. Farrell Lockheed Martn Advanced Technology
More informationCSCI 5417 Information Retrieval Systems Jim Martin!
CSCI 5417 Informaton Retreval Systems Jm Martn! Lecture 11 9/29/2011 Today 9/29 Classfcaton Naïve Bayes classfcaton Ungram LM 1 Where we are... Bascs of ad hoc retreval Indexng Term weghtng/scorng Cosne
More informationA mathematical programming approach to the analysis, design and scheduling of offshore oilfields
17 th European Symposum on Computer Aded Process Engneerng ESCAPE17 V. Plesu and P.S. Agach (Edtors) 2007 Elsever B.V. All rghts reserved. 1 A mathematcal programmng approach to the analyss, desgn and
More informationSupport Vector Machines
Support Vector Machnes Decson surface s a hyperplane (lne n 2D) n feature space (smlar to the Perceptron) Arguably, the most mportant recent dscovery n machne learnng In a nutshell: map the data to a predetermned
More informationAssembler. Building a Modern Computer From First Principles.
Assembler Buldng a Modern Computer From Frst Prncples www.nand2tetrs.org Elements of Computng Systems, Nsan & Schocken, MIT Press, www.nand2tetrs.org, Chapter 6: Assembler slde Where we are at: Human Thought
More informationA Background Subtraction for a Vision-based User Interface *
A Background Subtracton for a Vson-based User Interface * Dongpyo Hong and Woontack Woo KJIST U-VR Lab. {dhon wwoo}@kjst.ac.kr Abstract In ths paper, we propose a robust and effcent background subtracton
More informationNAG Fortran Library Chapter Introduction. G10 Smoothing in Statistics
Introducton G10 NAG Fortran Lbrary Chapter Introducton G10 Smoothng n Statstcs Contents 1 Scope of the Chapter... 2 2 Background to the Problems... 2 2.1 Smoothng Methods... 2 2.2 Smoothng Splnes and Regresson
More informationSRBIR: Semantic Region Based Image Retrieval by Extracting the Dominant Region and Semantic Learning
Journal of Computer Scence 7 (3): 400-408, 2011 ISSN 1549-3636 2011 Scence Publcatons SRBIR: Semantc Regon Based Image Retreval by Extractng the Domnant Regon and Semantc Learnng 1 I. Felc Raam and 2 S.
More informationSemi-Fragile Watermarking Scheme for Authentication of JPEG Images
Tamkang Journal of Scence and Engneerng, Vol. 10, No 1, pp. 5766 (2007) 57 Sem-Fragle Watermarkng Scheme for Authentcaton of JPEG Images Chh-Hung n 1 *, Tung-Shh Su 2 and Wen-Shyong Hseh 2,3 1 Department
More informationCAN COMPUTERS LEARN FASTER? Seyda Ertekin Computer Science & Engineering The Pennsylvania State University
CAN COMPUTERS LEARN FASTER? Seyda Ertekn Computer Scence & Engneerng The Pennsylvana State Unversty sertekn@cse.psu.edu ABSTRACT Ever snce computers were nvented, manknd wondered whether they mght be made
More informationOnline Detection and Classification of Moving Objects Using Progressively Improving Detectors
Onlne Detecton and Classfcaton of Movng Objects Usng Progressvely Improvng Detectors Omar Javed Saad Al Mubarak Shah Computer Vson Lab School of Computer Scence Unversty of Central Florda Orlando, FL 32816
More informationCollaboratively Regularized Nearest Points for Set Based Recognition
Academc Center for Computng and Meda Studes, Kyoto Unversty Collaboratvely Regularzed Nearest Ponts for Set Based Recognton Yang Wu, Mchhko Mnoh, Masayuk Mukunok Kyoto Unversty 9/1/013 BMVC 013 @ Brstol,
More informationEXTENDED BIC CRITERION FOR MODEL SELECTION
IDIAP RESEARCH REPORT EXTEDED BIC CRITERIO FOR ODEL SELECTIO Itshak Lapdot Andrew orrs IDIAP-RR-0-4 Dalle olle Insttute for Perceptual Artfcal Intellgence P.O.Box 59 artgny Valas Swtzerland phone +4 7
More informationAn Optimal Algorithm for Prufer Codes *
J. Software Engneerng & Applcatons, 2009, 2: 111-115 do:10.4236/jsea.2009.22016 Publshed Onlne July 2009 (www.scrp.org/journal/jsea) An Optmal Algorthm for Prufer Codes * Xaodong Wang 1, 2, Le Wang 3,
More informationImprovement of Spatial Resolution Using BlockMatching Based Motion Estimation and Frame. Integration
Improvement of Spatal Resoluton Usng BlockMatchng Based Moton Estmaton and Frame Integraton Danya Suga and Takayuk Hamamoto Graduate School of Engneerng, Tokyo Unversty of Scence, 6-3-1, Nuku, Katsuska-ku,
More informationSVM-based Learning for Multiple Model Estimation
SVM-based Learnng for Multple Model Estmaton Vladmr Cherkassky and Yunqan Ma Department of Electrcal and Computer Engneerng Unversty of Mnnesota Mnneapols, MN 55455 {cherkass,myq}@ece.umn.edu Abstract:
More informationLobachevsky State University of Nizhni Novgorod. Polyhedron. Quick Start Guide
Lobachevsky State Unversty of Nzhn Novgorod Polyhedron Quck Start Gude Nzhn Novgorod 2016 Contents Specfcaton of Polyhedron software... 3 Theoretcal background... 4 1. Interface of Polyhedron... 6 1.1.
More informationIntelligent Information Acquisition for Improved Clustering
Intellgent Informaton Acquston for Improved Clusterng Duy Vu Unversty of Texas at Austn duyvu@cs.utexas.edu Mkhal Blenko Mcrosoft Research mblenko@mcrosoft.com Prem Melvlle IBM T.J. Watson Research Center
More informationFeature Selection as an Improving Step for Decision Tree Construction
2009 Internatonal Conference on Machne Learnng and Computng IPCSIT vol.3 (2011) (2011) IACSIT Press, Sngapore Feature Selecton as an Improvng Step for Decson Tree Constructon Mahd Esmael 1, Fazekas Gabor
More informationFast Sparse Gaussian Processes Learning for Man-Made Structure Classification
Fast Sparse Gaussan Processes Learnng for Man-Made Structure Classfcaton Hang Zhou Insttute for Vson Systems Engneerng, Dept Elec. & Comp. Syst. Eng. PO Box 35, Monash Unversty, Clayton, VIC 3800, Australa
More informationAn Image Fusion Approach Based on Segmentation Region
Rong Wang, L-Qun Gao, Shu Yang, Yu-Hua Cha, and Yan-Chun Lu An Image Fuson Approach Based On Segmentaton Regon An Image Fuson Approach Based on Segmentaton Regon Rong Wang, L-Qun Gao, Shu Yang 3, Yu-Hua
More informationShape-adaptive DCT and Its Application in Region-based Image Coding
Internatonal Journal of Sgnal Processng, Image Processng and Pattern Recognton, pp.99-108 http://dx.do.org/10.14257/sp.2014.7.1.10 Shape-adaptve DCT and Its Applcaton n Regon-based Image Codng Yamn Zheng,
More informationProper Choice of Data Used for the Estimation of Datum Transformation Parameters
Proper Choce of Data Used for the Estmaton of Datum Transformaton Parameters Hakan S. KUTOGLU, Turkey Key words: Coordnate systems; transformaton; estmaton, relablty. SUMMARY Advances n technologes and
More informationEnhanced AMBTC for Image Compression using Block Classification and Interpolation
Internatonal Journal of Computer Applcatons (0975 8887) Volume 5 No.0, August 0 Enhanced AMBTC for Image Compresson usng Block Classfcaton and Interpolaton S. Vmala Dept. of Comp. Scence Mother Teresa
More informationKeywords - Wep page classification; bag of words model; topic model; hierarchical classification; Support Vector Machines
(IJCSIS) Internatonal Journal of Computer Scence and Informaton Securty, Herarchcal Web Page Classfcaton Based on a Topc Model and Neghborng Pages Integraton Wongkot Srura Phayung Meesad Choochart Haruechayasak
More informationMathematics 256 a course in differential equations for engineering students
Mathematcs 56 a course n dfferental equatons for engneerng students Chapter 5. More effcent methods of numercal soluton Euler s method s qute neffcent. Because the error s essentally proportonal to the
More informationLocal Quaternary Patterns and Feature Local Quaternary Patterns
Local Quaternary Patterns and Feature Local Quaternary Patterns Jayu Gu and Chengjun Lu The Department of Computer Scence, New Jersey Insttute of Technology, Newark, NJ 0102, USA Abstract - Ths paper presents
More informationAn Improved Image Segmentation Algorithm Based on the Otsu Method
3th ACIS Internatonal Conference on Software Engneerng, Artfcal Intellgence, Networkng arallel/dstrbuted Computng An Improved Image Segmentaton Algorthm Based on the Otsu Method Mengxng Huang, enjao Yu,
More informationModular PCA Face Recognition Based on Weighted Average
odern Appled Scence odular PCA Face Recognton Based on Weghted Average Chengmao Han (Correspondng author) Department of athematcs, Lny Normal Unversty Lny 76005, Chna E-mal: hanchengmao@163.com Abstract
More informationDistortion Function Designing for JPEG Steganography with Uncompressed Side-image
Dstorton Functon Desgnng for JPEG Steganography wth Uncompressed Sde-mage Fangjun Huang School of Informaton Scence and Technology, Sun Yat-Sen Unversty, GD 56, Chna huangfj@mal.sysu.edu.cn Jwu Huang School
More informationProblem Definitions and Evaluation Criteria for Computational Expensive Optimization
Problem efntons and Evaluaton Crtera for Computatonal Expensve Optmzaton B. Lu 1, Q. Chen and Q. Zhang 3, J. J. Lang 4, P. N. Suganthan, B. Y. Qu 6 1 epartment of Computng, Glyndwr Unversty, UK Faclty
More informationFace Recognition University at Buffalo CSE666 Lecture Slides Resources:
Face Recognton Unversty at Buffalo CSE666 Lecture Sldes Resources: http://www.face-rec.org/algorthms/ Overvew of face recognton algorthms Correlaton - Pxel based correspondence between two face mages Structural
More informationUB at GeoCLEF Department of Geography Abstract
UB at GeoCLEF 2006 Mguel E. Ruz (1), Stuart Shapro (2), June Abbas (1), Slva B. Southwck (1) and Davd Mark (3) State Unversty of New York at Buffalo (1) Department of Lbrary and Informaton Studes (2) Department
More informationConcurrent Apriori Data Mining Algorithms
Concurrent Apror Data Mnng Algorthms Vassl Halatchev Department of Electrcal Engneerng and Computer Scence York Unversty, Toronto October 8, 2015 Outlne Why t s mportant Introducton to Assocaton Rule Mnng
More informationLecture - Data Encryption Standard 4
The Data Encrypton Standard For an encrypton algorthm we requre: secrecy of the key and not of the algorthm tself s the only thng that s needed to ensure the prvacy of the data the best cryptographc algorthms
More informationMachine Learning. Support Vector Machines. (contains material adapted from talks by Constantin F. Aliferis & Ioannis Tsamardinos, and Martin Law)
Machne Learnng Support Vector Machnes (contans materal adapted from talks by Constantn F. Alfers & Ioanns Tsamardnos, and Martn Law) Bryan Pardo, Machne Learnng: EECS 349 Fall 2014 Support Vector Machnes
More informationMining Image Features in an Automatic Two- Dimensional Shape Recognition System
Internatonal Journal of Appled Mathematcs and Computer Scences Volume 2 Number 1 Mnng Image Features n an Automatc Two- Dmensonal Shape Recognton System R. A. Salam, M.A. Rodrgues Abstract The number of
More informationTECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS. Muradaliyev A.Z.
TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS Muradalyev AZ Azerbajan Scentfc-Research and Desgn-Prospectng Insttute of Energetc AZ1012, Ave HZardab-94 E-mal:aydn_murad@yahoocom Importance of
More informationHuman Face Recognition Using Generalized. Kernel Fisher Discriminant
Human Face Recognton Usng Generalzed Kernel Fsher Dscrmnant ng-yu Sun,2 De-Shuang Huang Ln Guo. Insttute of Intellgent Machnes, Chnese Academy of Scences, P.O.ox 30, Hefe, Anhu, Chna. 2. Department of
More informationSteganography is the art and science of hiding
Hde and Seek: An Introducton to Steganography Although people have hdden secrets n plan sght now called steganography throughout the ages the recent growth n computatonal power and technology has propelled
More informationFace Detection with Deep Learning
Face Detecton wth Deep Learnng Yu Shen Yus122@ucsd.edu A13227146 Kuan-We Chen kuc010@ucsd.edu A99045121 Yzhou Hao y3hao@ucsd.edu A98017773 Mn Hsuan Wu mhwu@ucsd.edu A92424998 Abstract The project here
More informationThree supervised learning methods on pen digits character recognition dataset
Three supervsed learnng methods on pen dgts character recognton dataset Chrs Flezach Department of Computer Scence and Engneerng Unversty of Calforna, San Dego San Dego, CA 92093 cflezac@cs.ucsd.edu Satoru
More information