Semantic Image Retrieval Using Region Based Inverted File

Size: px
Start display at page:

Download "Semantic Image Retrieval Using Region Based Inverted File"

Transcription

1 Semantc Image Retreval Usng Regon Based Inverted Fle Dengsheng Zhang, Md Monrul Islam, Guoun Lu and Jn Hou 2 Gppsland School of Informaton Technology, Monash Unversty Churchll, VIC 3842, Australa E-mal: { Dengsheng.Zhang, Md.Monrul.Islam,Guoun.Lu}@nfotech.monash.edu.au 2 School of Informaton Scence and Technology, Southwest Jaotong Unversty Chengdu, Schuan 6003, Chna Emal: hou@home.swtu.edu.cn Abstract Image data s as common as textual data n ths dgtal world. There s an urgent demand of mage management tools as effcent as those text search engnes. Decades of research on mage retreval has found there s a sgnfcant gap between the exstng content based mage retreval and semantc nterpretaton of mages by human. As a result, recent research on mage retreval has shfted to semantc mage retreval. Many semantc mage retreval models have been proposed, however, these methods are stll alenated from the wdely accepted text based retreval method. In ths paper, we propose to unte the semantc mage retreval model wth text based retreval usng a novel regon based nverted fle ndexng method. For ths purpose, mages are translated nto textual documents whch are then ndexed and retreved the same way as the conventonal text based search. Results show that our method not only provdes text based search effcency, but also better performance than the conventonal low level mage retreval. Keywords: CBIR, semantc mage retreval, nverted fle, decson tree. I. INTRODUCTION Wth the advancement of dgtal mage capturng devces and low cost electronc memory, huge amounts of mages are beng created everyday n dfferent areas. The need for the development of effcent and effectve methodologes to manage large mage databases for retreval s urgent. In the past, many content based mage retreval (CBIR) technques have been proposed [-2]. However, there s a sgnfcant gap between the low level mage features and hgh level semantcs. Recent research n mage retreval has shfted from CBIR to semantc based mage retreval, or SBIR for short. SBIR focuses on learnng semantcs from mage content or automatc mage annotaton [, 3]. The dea s to assocate the mage content wth semantcs usng machne learnng technques. Many SBIR technques have been proposed n lterature [4-3]. The representatve works n ths area nclude the co-occurrence model [4], cross-meda relevance model [5], translaton model [6], Gaussan mxture model [7-8], refnement model [9-0] and latent Drchlet allocaton model [-2]. Most of these technques are based on the classcal Bayesan theory. In these methods, a semantc model s learnt from a collecton of regon samples of each mage category. Once the semantc models are avalable, they are used for both mage annotaton and mage retreval. Gven an unknown mage, ts features are compared wth the learnt semantc models, the model wth the closest match to the mage s used as the annotaton of the mage. Gven a keyword query, mages n the database are ranked and retreved accordng to the probablty of annotatng each mage wth the query word. It can be seen that mage retreval s the same process as the annotaton, and the models are equvalent to the smlarty measure n the tradtonal CBIR. As the result, mage retreval n these cases s equvalent to the regon-regon matchng n the tradtonal regon based CBIR. Ths s mpractcal when the model s appled to large database, lke the Internet. At ths moment, semantc mage retreval s qute dfferent from textual document retreval, whch separates annotaton or keywords from documents and ndexes textual database wth keywords nstead of documents. Inspred by the textual document retreval technque, we adopt a dfferent approach from exstng semantc mage retreval methods. Specfcally, we frst buld a semantc dctonary and translate each mage document nto a set of textual keywords. We then ndex the mage database usng the nverted fle, so that mage retreval s done the same way as textual document retreval. However, the buldng of nverted fle for mage database s not straghorward. Unlke keywords n the textual documents, semantc keywords n mage documents carry dfferent weghts. Ths

2 s due to each semantc keyword s assocated wth regons of dfferent sze n each mage. Therefore, durng the ndexng, the mportance of each semantc keyword needs to be determned based on the regon sze the keyword s assocated to. Furthermore, spatal relatonshp between mage regons to whch the semantc keywords are assocated also need to be consdered durng the ndexng. The key dfference between our approach and exstng semantc mage retreval approach are n two aspects. Frst, we separate the semantc keywords from the mage documents and do mage annotaton and ndexng offlne. Second, we buld a regon based nverted fle to ndex mage documents the same way as textual documents, so that mage retreval s done the same way as textual document retreval. The rest of the paper s organzed as followng. In secton 2, we brefly descrbe how mage document s translated nto textual document. In secton 3, the proposed regon based nverted fle ndexng method s descrbed n detals. We present experment results n secton 4. The paper s concluded n secton 5. II. SEMANTIC LEARNING The basc dea of semantc learnng s to break down mages nto regons, learn semantcs from tranng regons and translate mages n database nto textual documents. Ths s llustrated n Fg..In the frst step, mages n the database are broken down nto regons usng automatc mage segmentaton algorthm such as JSEG tool [5]. Once mages are segmented nto regons, they are represented wth both colour and texture features, such as domnant colour descrptors (DCD) and Gabor texture features [3, 4]. After mages are segmented nto regons and represented usng colour and texture features, two vsual dctonares are created, they are, color dctonary and texture dctonary. A vsual dctonary s bascally a set of representatve feature vectors, t s analogue to a monolngual dctonary lke the Chnese dctonary or Englsh dctonary. It s generated from a set of tranng regons by clusterng the sample regons n the tranng sets. We have developed an adaptve vector quantzaton (AVQ) algorthm to generate the vsual dctonary. Once the vsual dctonares are generated, the colour feature vector of a regon s replaced by the ndex of the closest representatve feature (codeword) of the colour dctonary. A smlar approach s appled to dscretze the texture feature vectors. Once the vsual dctonares are created, a mappng between a semantc concept and codewords from dfferent vsual dctonares needs to be establshed. The set of all such mappngs forms the semantc dctonary whch s analogue to a blngual dctonary lke the Chnese-Englsh dctonary or Englsh-Chnese dctonary. In ths paper, we use decson tree (DT) to establsh these mappngs [3]. Regons can be better separated f both colour and texture dctonares are combned, therefore, both the color and texture dctonares are used n the semantc learnng. The annotaton usng DT conssts of two stages. The frst stage s the tranng stage where the DT s traned by labellng the regons n the tranng dataset whch s obtaned from ground truth. The DT s bult upon the basc C4.5 algorthm wth an addtonal prunng process. C4.5 has been chosen because the attrbute selecton process of C4.5 s proved to better than that of ID3 [3]. Colour and texture attrbutes are frst converted to nomnal values usng the vsual dctonares. Next, the nomnal values of the tranng sets are nput to the DT for nducton. Once the DT s bult and traned wth the tranng set, t s used as the semantc dctonary to annotate unknown regons of an mage n the second stage. After the annotaton, each mage n database s translated nto a set of keywords for ndexng. Sky Bush Sky Learnng Anmal Annotaton House Horse Fgure. Block dagram of translatng mage nto textual documents.

3 III. SEMANTIC IMAGE INDEXING AND RETRIEVAL USING INVERTED FILE Once the mages are translated nto textual documents, they can be ndexed and retreved usng the effcent nverted fle technque. In ths secton, we fst brefly descrbe the conventonal text based nverted fle ndexng and retreval technque. We then descrbe the proposed regon based nverted fle for semantc mage ndexng and retreval. A. Text Based Inverted Fle An nverted fle s a data structure where documents are ndexed by term documents structure nstead of the conventonal document term structure. Typcally, an nverted fle s a collecton of lsts - one lst for each term. To form an nverted fle, all the documents of a database are parsed to fnd the canddate keywords or terms. Suppose the total numbers of documents and terms are N and M (M<<N), respectvely. The nverted fle contans one lst for each term. For each term, term, the followng nformaton are collected and stored wth the term n the correspondng lst: Document frequency df : the number of documents contanng the term term Document ID doc : ndcatng whch document contans the term term Term frequency : the number of tmes term appears n the th document Table I shows an nverted fle for textual documents. It s understood that f a term appears frequently n a document, the term s hghly relevant to that document. Therefore, ensures that hgh frequency terms have more weght than low frequency terms. However, f a term appears n many documents, ths term has lttle relevance to any partcular document, and ths term should have less weght than others. Ths s defne by the nverse document frequency df and s calculated as, df = N log( ) df () and df are used to calculate the weght (mportance) of term n the th document, tw = df (2), The th document s then represented wth an M dmensonal feature vector, F = tw,, tw,2, tw,3, L, tw, M (3) The query text s also converted nto feature vector of dmenson M, Q = q, q2, q 3, L, q M (4) where, q takes bnary values and s defned as,, f term appears n the query text Q q = (5) 0, otherwse Terms Document Frequency TABLE I: INVERTED FILE FOR TEXTUAL DOCUMENTS. Inverted Lst (Documents & term frequency ) term df 2 df < doc, >, < doc2, >, L, < doc df, > M M M term df 2 df < doc, >, < doc2, >, L, < docdf, > M M M 2 df term M df M M < doc, M >, < doc2, M >, L, < docdf, > M M

4 The smlarty between a query text Q and the th document D s measured as, Smlarty( D, Q) = ( tw, q ) M 2 tw, q M M The documents are then ranked accordng to the smlarty scores and returned to users. B. Image Indexng Usng Inverted Fle Snce mages have been translated nto textual documents, mages can be ndexed and retreved the same way as the textual document ndexng and retreval usng nverted fle data structure. However, due to the dfference between textual and vsual words, the term frequency and term weght of the textual documents need to be redesgned for mages. Furthermore, addtonal nformaton lke spatal nformaton of mages should also be ncluded n the nverted fle to ncrease retreval accuracy. In textual documents, each occurrence of a term n a document gves the same nformaton as other occurrence of the same term, therefore the weght of a term s well determned by the term frequency. For mages, however, each occurrence of a term (concept) may represent a regon wth dfferent sze from another occurrence of the same term. For example, n Fgure 2, three regons wth dfferent szes belong to the same concept flower. If the term frequency 3 were used, t would not capture the mportance of the large flower n the centre, because another mage may have a much smaller background flower also wth 3 segmented regons. Therefore, nstead of usng ust the term frequency, the regons assocated wth the same term are summed together and a term area s calculated to represent the mportance of the term flower. Ths term area s the term weght of flower n ths partcular mage. As the result, the term weght tw, redefned as the sum of areas of all regons assocated wth the term n the th mage. In addton to term frequency, there are other dfferences between mages and text documents, whch affect the calculaton of the term weght. Images consst of (6) s pxels n 2D coordnate system, they contan rch spatal nformaton. For example, an anmal usually appears n the mddle, water n the bottom, sky n the upper part of the mages as shown n Fgure 3. If a regon s annotated as an anmal regon, the accuracy of the retreval can be mproved f the regon s found n the centre of the mage. Therefore, an anmal regon appearng n the centre should have more weght than an anmal regon appearng elsewhere n the mage. The poston weght of a regon reg k n the th mage document s calculated usng Equaton (7) as followng: d pos _ weghtk = 2 ( ) (7) dmax Fgure 2: Multple regons wth flower concept. Fgure 3: Examples of spatal postons. Sky n the top, water n the bottom and anmals n the centre. where, d s the dstance of the regon centrod from the centre of ts normal poston n an mage and d max s ts maxmum possble dstance. Fgure 4 shows typcal examples of how to calculate d and d max for an anmal regon, a cloud regon and a water regon. Equaton (7) ensures that the farther s the regon from the centre, the less s the poston weght. d and d max for other concepts are calculated n a smlar way. In TABLE II, we defne how d and d max are calculated for 3 concepts used n ths paper.

5 d d d max d max d d max (a) (b) (c) Fgure 4: Calculaton of d and d max for (a) anmal, (b) sky and (c) grass regons. TABLE II: DEFINITIONS OF d AND d max FOR DIFFERENT CONCEPTS. Concept of the regon Assumpton of normal poston d Dstance between d max Dstance between anmal, car, frut, flower Centre Image centrod and regon centrod Image centrod and furthest corner frework, sky, brd Top Top centre and regon centrod Top centre and furthest corner grass, sand, water Bottom Bottom centre and regon centrod Bottom centre and furthest corner Another mportant pror nformaton s the co-occurrence of obects n many mage categores. In mages, certan obects often appear concurrently, for example tger and grass, brd and sky, beach and sea, etc, as shown n Fgure 5. If a regon s labelled as a brd regon, the accuracy of retreval can be mproved f t appears wth sky regon n the same mage. Therefore, the weght of a brd regon s ncreased f t appears wth a sky regon. Ths dea s mplemented by ntroducng relatonshp weght of a regon, rel _ weght k whch s calculated as the frequency of co-occurrng regons for the regon reg k n the th mage. For example, f a regon s labelled as brd and two other regons are labelled as sky n the same mage, the relatonshp weght of brd regon wll be 3 (=+2). We have studed the mage database and have made a lst of cooccurrng obects. TABLE III shows some of co-occurrng obects used n the semantc retreval. Because of the above mentoned addtonal nformaton, an nverted fle for mage documents needs to store more nformaton than an nverted fle for conventonal textual documents. For example, area, pos_weght and rel_weght are stored for each regon. TABLE IV shows the structure of an nverted fle for mage documents. Analogous to TABLE I, each concept term s assocated wth followng set of nformaton: Document frequency df : the number of mages contanng one or more regons labelled wth term (or concept ) Image ID m : ndcatng the mage whch contans one or more regons labelled wth term Term frequency wth term n the th mage Regon nformaton lst number of regon records as : the number of regons labelled reginfo : contanng {( reginfo = a, p, r ),( a2, p2, r 2 ) L,( a, p, r where, a, p and r are respectvely the area, pos_weght and rel_weght of a regon. The nformaton from Table IV s used to compute tw, whch s done n offlne durng actual ndexng: tw = df ( areak pos weghtk rel _ weght ) k k= Inverse document frequency )} _ (8) df s calculated n a smlar way of Equaton (). The modfed term weght tw of (8) replaces Equaton (3) to extract the feature vector for each mage. All mages are represented and

6 ndexed usng these feature vectors. Durng retreval, a query text s gven wth one or more keywords. The query text s converted nto a query feature vector as shown n Equaton (4). The smlartes between query feature vector and all feature vectors of database mage are calculated usng Equaton (6). Images are ranked accordng to the smlarty values and are dsplayed to the users. Fgure 5: Examples of co-occurrng obects n mages. Top: sky and brd. Bottom: tger and grass. Concept Name Black Ape Black Elephant Brown Horse Buldng Butterfly Camel Caulflower Corn Deer Eagle Eggplant Fghter Plane Golden Fsh Greyhound Dog House TABLE III: LIST OF CO-OCCURRING CONCEPTS. Frequently occurrng concepts Forest, Brown, Green Brown, Green, Black Water, Blue Water Tree, Forest, Brown, Green Blue Sky, Cloudy Sky, Other Sky Tree, Forest, Brown, Green, Flower Green, Sand Green, Green Leaf Green, Green Leaf Brown, Green Blue Sky, Cloudy Sky, Other Sky Green, Green Leaf Blue Sky, Cloudy Sky, Other Sky Black Water, Blue Water Green, Green Leaf, Buldng, House Blue Sky, Cloudy Sky, Other Sky Term Image Freq. term df TABLE IV: INVERTED FILE FOR IMAGE DOCUMENTS. Inverted Lsts 2 2 df df m,, reginfo, m2,, reginfo, L, mdf,, reginfo M M M 2 2 df df term df m,, reginfo, m2,, reginfo, L, mdf,, reginfo M M M 2 2 df term M df M df M m, M, reginfom, m2, M, reginfom, L, mdf, M, reginfom M M IV. EXPERIMENTAL RESULTS In ths secton, we test the performance of semantc ndexng and retreval usng nverted fle. The DT s traned usng a set of labelled regons. We collect 7,600 mages. 5,00 mages are from Corel dataset. The remanng 2,500 mages are downloaded from the Internet by searchng Google and Yahoo wth 50 dfferent keywords. All mages are segmented nto regons usng JSEG segmentaton algorthm. In total, 55,870 regons are obtaned. Each regon s represented wth a colour descrptor and a texture descrptor. Domnant colour descrptor (DCD) s used as the colour descrptor, Gabor features are used as the texture descrptor. In total, there are 02 concepts n the mage database.

7 To tran and test the DT, we select 6,088 regons belongng to dfferent concepts. The number of regons s dfferent for each concept. For each concept, we randomly select 50% regons for tranng and other 50% regons for testng. Ths process generates a tranng set and a testng set both consstng of 3,044 regons. Each regon n the tranng and testng set s descrbed by two attrbutes (color and texture) and a class label. All the attrbutes are multdmensonal vectors. Among them, the colour attrbute has varable dmenson and the texture attrbute has fxed dmensons. They are dscretzed usng the AVQ algorthm and fed nto the DT algorthm descrbed n Secton 2. Wth the DT annotaton, we are able to acheve 67% annotaton accuracy compared wth 5% accuracy by the wdely used Bayesan method n lterature (Fg. 6(a)). We then compare semantc mage retreval wth low level mage retreval. In low level retreval, mages are segmented nto regons the same way as above. However, nstead of keyword based representaton, each regon n an mage s represented by low level colour and texture features. To be far, all the colour and texture features used n the annotaton stage are used to represent each regon. Images are ndexed as { mage regon} and mage retreval s done by mage by mage matchng. For low level retreval, 02 mages are selected as queres - one query mage for each concept. We select those query mages whch gve the best low level retreval performance. As the number of regons n dfferent mages s dfferent, we use wdely used earth mover s dstance (EMD) to fnd the dstance between two mages [3]. Gven a query mage, all database mages are ranked based on ther EMD dstance from the query. The retreval accuracy s reported usng precson on top k retreved mages, where k vares by 0, 20, 30,..., 00. For each query, precson s calculated at each level of k. Precsons of all queres are then averaged. For semantc retreval, we have 02 dfferent queres representng 02 concepts. The retreval precsons are averaged usng 02 queres. Fg. 6(b) shows the comparson of the two retreval methods. The sgnfcant dfference between the two types of retreval demonstrates the effectveness of learnng. In low level retreval, due to feature naccuracy, smlar mages are often placed far way and dfferent mages are also often placed close to each other n hgh dmensonal feature space. Therefore, a low level retreval often results n many rrelevant mages. Ths problem s overcome n SBIR by learnng whch ncorporate human knowledge. For example, two smlar mages may be placed far away each other, n SBIR, however they are placed n the same cluster due to learnng. Therefore, they are regarded as relevant mages n SBIR. As a result the retreval performance ncreases. Precson (%) Annotaton Accuracy (%) DDT (a) Bayesan Top K retreved mages (b) Semantcl retreval Low level retreval Fgure 6. (a) comparson of annotaton accuracy; (b) comparson of retreval accuracy.

8 V. CONCLUSIONS We have proposed a method to translate mages nto textual documents and ndex mages usng a regon based nverted fle. We demonstrate that there s a natural relatonshp between mage data and textual data. The novelty of our approach s that nstead of dong annotaton onlne lke those exstng methods n lterature, we separate semantcs from mage documents and ndex mages wth semantc terms n offlne. Ths s equvalent to classfy mages n database nto semantc classes and then retreve mages class by class. Ths s much more effcent than exstng methods when the retreval model s appled to large scale database lke the Internet. Because t s mpractcal n ths case to navgate through all the mages n the database, rather, t s only practcal to select mages relevant to the query for rankng. Another novelty of our approach s the ncluson of some of the mportant pror nformaton n the nverted fle ndexng. Results show our semantc learnng has an advantage over the wdely used Bayesan annotaton method n lterature and has hgher retreval performance than the tradtonal CBIR method. In future, we wll nvestgate the varous metadata sources and combne mage content data wth metadata f they are avalable. The current concept set wll also be expanded to adapt the system to larger databases. REFERENCES [] R. Datta, D. Josh, J. L and J. Z. Wang, Image retreval: deas, nfluences, and trends of the new age, ACM Computng Surveys, 40(2):5:-60, [2] F. Long, H. J. Zhang and D. D. Feng, Fundamentals of content-based mage retreval, Multmeda Informaton Retreval and Management, D. Feng Eds, Sprnger, [3] Y. Lu, D. S. Zhang and G. Lu, "A survey of contentbased mage retreval wth hgh-level semantcs", Pattern Recognton, 40(): , [4] Y. Mor, H. Takahash and R. Oka, Image-to-word transformaton based on dvdng and vector quantzng mages wth words, Proc. of Internatonal. Workshop on Multmeda Intellgent Storage and Retreval Management, 999. [5] J. Jeon, V. Lavrenko and R. Manmatha, Automatc mage annotaton and retreval usng cross-meda relevance models, Proc. of ACM SIGIR03, pp.9 26, [6] P. Duygulu, K. Barnard, J. F. G. de. Fretas, and D. A. Forsyth, Obect recognton as machne translaton: learnng a lexcon for a fxed mage vocabulary, Proc. of ECCV02, pp.97-2, [7] J. L and J. Z. Wang, Real-tme computerzed annotaton of pctures, IEEE PAMI, 30(6): , [8] G. Carnero A. B. Chan, P. J. Moreno and N. Vasconcelos, Supervsed learnng of semantc classes for mage annotaton and retreval, IEEE PAMI, 29(3):394-40, [9] Y. Lu, D. Zhang and G. Lu, Seve-search mages effectvely through vsual elmnaton, Lecture Notes n Computer Scence, 4577:38-390, [0] C. Wang, F. Jng, L. Zhang and H.-J. Zhang, Content-based mage annotaton refnement, Proc. of CVPR, [] L. Fe-Fe and P. Perona, A Bayesan herarchy model for learnng natural scene categores, Proc. of CVPR, [2] D. Ble, A. Ng and M. Jordan, Latent Drchlet allocaton, Journal of Machne Learnng Research, 3: , [3] Y. Lu, D. Zhang and G. Lu, Regon-based mage retreval wth hgh-level semantcs usng decson tree learnng, Pattern Recognton, 4(8): , [4] D. Zhang, A. Wong, M. Indrawan and G. Lu. "Content-based mage retreval usng gabor texture features", Proc. of Frst IEEE Pacfc-Rm Conference on Multmeda (PCM'00), pp , Sydney, Australa, [5] Y. Deng and B. S. Manunath, "Unsupervsed Segmentaton of Color-Texture Regons n Images and Vdeo," IEEE Trans. on Pattern Analyss and Machne Learnng (PAMI), 23(8):800-80, 200.

A Fast Content-Based Multimedia Retrieval Technique Using Compressed Data

A 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 information

Content Based Image Retrieval Using 2-D Discrete Wavelet with Texture Feature with Different Classifiers

Content 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 information

Term Weighting Classification System Using the Chi-square Statistic for the Classification Subtask at NTCIR-6 Patent Retrieval Task

Term 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 information

A Binarization Algorithm specialized on Document Images and Photos

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 information

Machine Learning: Algorithms and Applications

Machine 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 information

A Unified Framework for Semantics and Feature Based Relevance Feedback in Image Retrieval Systems

A 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 information

A PATTERN RECOGNITION APPROACH TO IMAGE SEGMENTATION

A 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 information

Learning the Kernel Parameters in Kernel Minimum Distance Classifier

Learning 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 information

A Fast Visual Tracking Algorithm Based on Circle Pixels Matching

A 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 information

Machine Learning 9. week

Machine Learning 9. week Machne Learnng 9. week Mappng Concept Radal Bass Functons (RBF) RBF Networks 1 Mappng It s probably the best scenaro for the classfcaton of two dataset s to separate them lnearly. As you see n the below

More information

Improvement of Spatial Resolution Using BlockMatching Based Motion Estimation and Frame. Integration

Improvement 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 information

Improving Web Image Search using Meta Re-rankers

Improving Web Image Search using Meta Re-rankers VOLUME-1, ISSUE-V (Aug-Sep 2013) IS NOW AVAILABLE AT: www.dcst.com Improvng Web Image Search usng Meta Re-rankers B.Kavtha 1, N. Suata 2 1 Department of Computer Scence and Engneerng, Chtanya Bharath Insttute

More information

Relevance Feedback for Image Retrieval

Relevance Feedback for Image Retrieval Vashal D Dhale et al, / (IJCSIT Internatonal Journal of Computer Scence and Informaton Technologes, Vol 4 (2, 203, 39-323 Relevance Feedback for Image Retreval Vashal D Dhale, Dr A R Mahaan, Prof Uma Thakur

More information

SRBIR: Semantic Region Based Image Retrieval by Extracting the Dominant Region and Semantic Learning

SRBIR: 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 information

Steps for Computing the Dissimilarity, Entropy, Herfindahl-Hirschman and. Accessibility (Gravity with Competition) Indices

Steps for Computing the Dissimilarity, Entropy, Herfindahl-Hirschman and. Accessibility (Gravity with Competition) Indices Steps for Computng the Dssmlarty, Entropy, Herfndahl-Hrschman and Accessblty (Gravty wth Competton) Indces I. Dssmlarty Index Measurement: The followng formula can be used to measure the evenness between

More information

SLAM Summer School 2006 Practical 2: SLAM using Monocular Vision

SLAM 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 information

FEATURE EXTRACTION. Dr. K.Vijayarekha. Associate Dean School of Electrical and Electronics Engineering SASTRA University, Thanjavur

FEATURE 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 information

MULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION

MULTISPECTRAL 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 information

Visual Thesaurus for Color Image Retrieval using Self-Organizing Maps

Visual Thesaurus for Color Image Retrieval using Self-Organizing Maps Vsual Thesaurus for Color Image Retreval usng Self-Organzng Maps Chrstopher C. Yang and Mlo K. Yp Department of System Engneerng and Engneerng Management The Chnese Unversty of Hong Kong, Hong Kong ABSTRACT

More information

Object-Based Techniques for Image Retrieval

Object-Based Techniques for Image Retrieval 54 Zhang, Gao, & Luo Chapter VII Object-Based Technques for Image Retreval Y. J. Zhang, Tsnghua Unversty, Chna Y. Y. Gao, Tsnghua Unversty, Chna Y. Luo, Tsnghua Unversty, Chna ABSTRACT To overcome the

More information

Fuzzy C-Means Initialized by Fixed Threshold Clustering for Improving Image Retrieval

Fuzzy C-Means Initialized by Fixed Threshold Clustering for Improving Image Retrieval Fuzzy -Means Intalzed by Fxed Threshold lusterng for Improvng Image Retreval NAWARA HANSIRI, SIRIPORN SUPRATID,HOM KIMPAN 3 Faculty of Informaton Technology Rangst Unversty Muang-Ake, Paholyotn Road, Patumtan,

More information

Description of NTU Approach to NTCIR3 Multilingual Information Retrieval

Description of NTU Approach to NTCIR3 Multilingual Information Retrieval Proceedngs of the Thrd NTCIR Workshop Descrpton of NTU Approach to NTCIR3 Multlngual Informaton Retreval Wen-Cheng Ln and Hsn-Hs Chen Department of Computer Scence and Informaton Engneerng Natonal Tawan

More information

COMPLEX WAVELET TRANSFORM-BASED COLOR INDEXING FOR CONTENT-BASED IMAGE RETRIEVAL

COMPLEX WAVELET TRANSFORM-BASED COLOR INDEXING FOR CONTENT-BASED IMAGE RETRIEVAL COMPLEX WAVELET TRANSFORM-BASED COLOR INDEXING FOR CONTENT-BASED IMAGE RETRIEVAL Nader Safavan and Shohreh Kasae Department of Computer Engneerng Sharf Unversty of Technology Tehran, Iran skasae@sharf.edu

More information

Background Removal in Image indexing and Retrieval

Background Removal in Image indexing and Retrieval Background Removal n Image ndexng and Retreval Y Lu and Hong Guo Department of Electrcal and Computer Engneerng The Unversty of Mchgan-Dearborn Dearborn Mchgan 4818-1491, U.S.A. Voce: 313-593-508, Fax:

More information

Detection of an Object by using Principal Component Analysis

Detection 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 information

An Image Fusion Approach Based on Segmentation Region

An 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 information

Parallelism for Nested Loops with Non-uniform and Flow Dependences

Parallelism for Nested Loops with Non-uniform and Flow Dependences Parallelsm for Nested Loops wth Non-unform and Flow Dependences Sam-Jn Jeong Dept. of Informaton & Communcaton Engneerng, Cheonan Unversty, 5, Anseo-dong, Cheonan, Chungnam, 330-80, Korea. seong@cheonan.ac.kr

More information

Pruning Training Corpus to Speedup Text Classification 1

Pruning Training Corpus to Speedup Text Classification 1 Prunng Tranng Corpus to Speedup Text Classfcaton Jhong Guan and Shugeng Zhou School of Computer Scence, Wuhan Unversty, Wuhan, 430079, Chna hguan@wtusm.edu.cn State Key Lab of Software Engneerng, Wuhan

More information

A Novel Adaptive Descriptor Algorithm for Ternary Pattern Textures

A Novel Adaptive Descriptor Algorithm for Ternary Pattern Textures A Novel Adaptve Descrptor Algorthm for Ternary Pattern Textures Fahuan Hu 1,2, Guopng Lu 1 *, Zengwen Dong 1 1.School of Mechancal & Electrcal Engneerng, Nanchang Unversty, Nanchang, 330031, Chna; 2. School

More information

Outline. Discriminative classifiers for image recognition. Where in the World? A nearest neighbor recognition example 4/14/2011. CS 376 Lecture 22 1

Outline. 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 information

Lobachevsky State University of Nizhni Novgorod. Polyhedron. Quick Start Guide

Lobachevsky 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 information

1. Introduction. Abstract

1. Introduction. Abstract Image Retreval Usng a Herarchy of Clusters Danela Stan & Ishwar K. Seth Intellgent Informaton Engneerng Laboratory, Department of Computer Scence & Engneerng, Oaland Unversty, Rochester, Mchgan 48309-4478

More information

An Optimal Algorithm for Prufer Codes *

An 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 information

Query Clustering Using a Hybrid Query Similarity Measure

Query Clustering Using a Hybrid Query Similarity Measure Query clusterng usng a hybrd query smlarty measure Fu. L., Goh, D.H., & Foo, S. (2004). WSEAS Transacton on Computers, 3(3), 700-705. Query Clusterng Usng a Hybrd Query Smlarty Measure Ln Fu, Don Hoe-Lan

More information

Combination of Color and Local Patterns as a Feature Vector for CBIR

Combination of Color and Local Patterns as a Feature Vector for CBIR Internatonal Journal of Computer Applcatons (975 8887) Volume 99 No.1, August 214 Combnaton of Color and Local Patterns as a Feature Vector for CBIR L.Koteswara Rao Asst.Professor, Dept of ECE Faculty

More information

Support Vector Machines

Support 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 information

Cluster Analysis of Electrical Behavior

Cluster 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 information

Edge Detection in Noisy Images Using the Support Vector Machines

Edge 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 information

User Authentication Based On Behavioral Mouse Dynamics Biometrics

User 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 information

Color Image Segmentation Using Multispectral Random Field Texture Model & Color Content Features

Color Image Segmentation Using Multispectral Random Field Texture Model & Color Content Features Color Image Segmentaton Usng Multspectral Random Feld Texture Model & Color Content Features Orlando J. Hernandez E-mal: hernande@tcnj.edu Department Electrcal & Computer Engneerng, The College of New

More information

Subspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points;

Subspace 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 information

A Clustering Algorithm for Key Frame Extraction Based on Density Peak

A Clustering Algorithm for Key Frame Extraction Based on Density Peak Journal of Computer and Communcatons, 2018, 6, 118-128 http://www.scrp.org/ournal/cc ISSN Onlne: 2327-5227 ISSN Prnt: 2327-5219 A Clusterng Algorthm for Key Frame Extracton Based on Densty Peak Hong Zhao

More information

A New Feature of Uniformity of Image Texture Directions Coinciding with the Human Eyes Perception 1

A New Feature of Uniformity of Image Texture Directions Coinciding with the Human Eyes Perception 1 A New Feature of Unformty of Image Texture Drectons Concdng wth the Human Eyes Percepton Xng-Jan He, De-Shuang Huang, Yue Zhang, Tat-Mng Lo 2, and Mchael R. Lyu 3 Intellgent Computng Lab, Insttute of Intellgent

More information

KIDS Lab at ImageCLEF 2012 Personal Photo Retrieval

KIDS Lab at ImageCLEF 2012 Personal Photo Retrieval KD Lab at mageclef 2012 Personal Photo Retreval Cha-We Ku, Been-Chan Chen, Guan-Bn Chen, L-J Gaou, Rong-ng Huang, and ao-en Wang Knowledge, nformaton, and Database ystem Laboratory Department of Computer

More information

Cross-Language Information Retrieval

Cross-Language Information Retrieval Feature Artcle: Cross-Language Informaton Retreval 19 Cross-Language Informaton Retreval Jan-Yun Ne 1 Abstract A research group n Unversty of Montreal has worked on the problem of cross-language nformaton

More information

Tsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance

Tsinghua 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 information

HIGH-LEVEL SEMANTICS OF IMAGES IN WEB DOCUMENTS USING WEIGHTED TAGS AND STRENGTH MATRIX

HIGH-LEVEL SEMANTICS OF IMAGES IN WEB DOCUMENTS USING WEIGHTED TAGS AND STRENGTH MATRIX HIGH-LEVEL SEMANTICS OF IMAGES IN WEB DOCUMENTS USING WEIGHTED TAGS AND STRENGTH MATRIX P.Shanmugavadvu 1, P.Sumathy 2, A.Vadvel 3 12 Department of Computer Scence and Applcatons, Gandhgram Rural Insttute,

More information

Semantic Illustration Retrieval for Very Large Data Set

Semantic Illustration Retrieval for Very Large Data Set Semantc Illustraton Retreval for Very Large Data Set Song Ka, Huang Te-Jun, Tan Yong-Hong Dgtal Meda Lab, Insttute of Computng Technology, Chnese Academy of Scences Beng, 00080, R Chna Insttute for Dgtal

More information

Can We Beat the Prefix Filtering? An Adaptive Framework for Similarity Join and Search

Can We Beat the Prefix Filtering? An Adaptive Framework for Similarity Join and Search Can We Beat the Prefx Flterng? An Adaptve Framework for Smlarty Jon and Search Jannan Wang Guolang L Janhua Feng Department of Computer Scence and Technology, Tsnghua Natonal Laboratory for Informaton

More information

Semantic Scene Concept Learning by an Autonomous Agent

Semantic Scene Concept Learning by an Autonomous Agent Semantc Scene Concept Learnng by an Autonomous Agent Weyu Zhu Illnos Wesleyan Unversty PO Box 29, Bloomngton, IL 672 wzhu@wu.edu Abstract Scene understandng addresses the ssue of what a scene contans.

More information

Using Fuzzy Logic to Enhance the Large Size Remote Sensing Images

Using Fuzzy Logic to Enhance the Large Size Remote Sensing Images Internatonal Journal of Informaton and Electroncs Engneerng Vol. 5 No. 6 November 015 Usng Fuzzy Logc to Enhance the Large Sze Remote Sensng Images Trung Nguyen Tu Huy Ngo Hoang and Thoa Vu Van Abstract

More information

Learning-Based Top-N Selection Query Evaluation over Relational Databases

Learning-Based Top-N Selection Query Evaluation over Relational Databases Learnng-Based Top-N Selecton Query Evaluaton over Relatonal Databases Lang Zhu *, Wey Meng ** * School of Mathematcs and Computer Scence, Hebe Unversty, Baodng, Hebe 071002, Chna, zhu@mal.hbu.edu.cn **

More information

A Feature-Weighted Instance-Based Learner for Deep Web Search Interface Identification

A Feature-Weighted Instance-Based Learner for Deep Web Search Interface Identification Research Journal of Appled Scences, Engneerng and Technology 5(4): 1278-1283, 2013 ISSN: 2040-7459; e-issn: 2040-7467 Maxwell Scentfc Organzaton, 2013 Submtted: June 28, 2012 Accepted: August 08, 2012

More information

Corner-Based Image Alignment using Pyramid Structure with Gradient Vector Similarity

Corner-Based Image Alignment using Pyramid Structure with Gradient Vector Similarity Journal of Sgnal and Informaton Processng, 013, 4, 114-119 do:10.436/jsp.013.43b00 Publshed Onlne August 013 (http://www.scrp.org/journal/jsp) Corner-Based Image Algnment usng Pyramd Structure wth Gradent

More information

Hierarchical Image Retrieval by Multi-Feature Fusion

Hierarchical Image Retrieval by Multi-Feature Fusion Preprnts (www.preprnts.org) NOT PEER-REVIEWED Posted: 26 Aprl 207 do:0.20944/preprnts20704.074.v Artcle Herarchcal Image Retreval by Mult- Fuson Xaojun Lu, Jaojuan Wang,Yngq Hou, Me Yang, Q Wang* and Xangde

More information

Efficient Segmentation and Classification of Remote Sensing Image Using Local Self Similarity

Efficient Segmentation and Classification of Remote Sensing Image Using Local Self Similarity ISSN(Onlne): 2320-9801 ISSN (Prnt): 2320-9798 Internatonal Journal of Innovatve Research n Computer and Communcaton Engneerng (An ISO 3297: 2007 Certfed Organzaton) Vol.2, Specal Issue 1, March 2014 Proceedngs

More information

UB at GeoCLEF Department of Geography Abstract

UB 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 information

Performance Evaluation of Information Retrieval Systems

Performance Evaluation of Information Retrieval Systems Why System Evaluaton? Performance Evaluaton of Informaton Retreval Systems Many sldes n ths secton are adapted from Prof. Joydeep Ghosh (UT ECE) who n turn adapted them from Prof. Dk Lee (Unv. of Scence

More information

A Novel Video Retrieval Method Based on Web Community Extraction Using Features of Video Materials

A Novel Video Retrieval Method Based on Web Community Extraction Using Features of Video Materials IEICE TRANS. FUNDAMENTALS, VOL.E92 A, NO.8 AUGUST 2009 1961 PAPER Specal Secton on Sgnal Processng A Novel Vdeo Retreval Method Based on Web Communty Extracton Usng Features of Vdeo Materals Yasutaka HATAKEYAMA

More information

Classifier Selection Based on Data Complexity Measures *

Classifier 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 information

A CALCULATION METHOD OF DEEP WEB ENTITIES RECOGNITION

A CALCULATION METHOD OF DEEP WEB ENTITIES RECOGNITION A CALCULATION METHOD OF DEEP WEB ENTITIES RECOGNITION 1 FENG YONG, DANG XIAO-WAN, 3 XU HONG-YAN School of Informaton, Laonng Unversty, Shenyang Laonng E-mal: 1 fyxuhy@163.com, dangxaowan@163.com, 3 xuhongyan_lndx@163.com

More information

Arabic Text Classification Using N-Gram Frequency Statistics A Comparative Study

Arabic Text Classification Using N-Gram Frequency Statistics A Comparative Study Arabc Text Classfcaton Usng N-Gram Frequency Statstcs A Comparatve Study Lala Khresat Dept. of Computer Scence, Math and Physcs Farlegh Dcknson Unversty 285 Madson Ave, Madson NJ 07940 Khresat@fdu.edu

More information

Searching Large Image Databases using Color Information

Searching Large Image Databases using Color Information Searchng Large Image Databases usng Color Informaton Ioan Racu CMSC350: Artfcal Intellgence Wnter Quarter 2004 Department of Computer Scence Unversty of Chcago racu@cs.uchcago.edu ABSTRACT The goal of

More information

CSCI 5417 Information Retrieval Systems Jim Martin!

CSCI 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 information

A Gradient Difference based Technique for Video Text Detection

A Gradient Difference based Technique for Video Text Detection A Gradent Dfference based Technque for Vdeo Text Detecton Palaahnakote Shvakumara, Trung Quy Phan and Chew Lm Tan School of Computng, Natonal Unversty of Sngapore {shva, phanquyt, tancl }@comp.nus.edu.sg

More information

Local Quaternary Patterns and Feature Local Quaternary Patterns

Local 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 information

Multi-modal Multi-label Semantic Indexing of Images Based on Hybrid Ensemble Learning

Multi-modal Multi-label Semantic Indexing of Images Based on Hybrid Ensemble Learning Mult-modal Mult-label Semantc Indexng of Images Based on Hybrd Ensemble Learnng We L 1, Maosong Sun 1, and Chrstopher Habel 2 1 State Key Lab of Intellgent Technology and Systems Department of Computer

More information

Skew Angle Estimation and Correction of Hand Written, Textual and Large areas of Non-Textual Document Images: A Novel Approach

Skew 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 information

A Gradient Difference based Technique for Video Text Detection

A Gradient Difference based Technique for Video Text Detection 2009 10th Internatonal Conference on Document Analyss and Recognton A Gradent Dfference based Technque for Vdeo Text Detecton Palaahnakote Shvakumara, Trung Quy Phan and Chew Lm Tan School of Computng,

More information

Novel Pattern-based Fingerprint Recognition Technique Using 2D Wavelet Decomposition

Novel Pattern-based Fingerprint Recognition Technique Using 2D Wavelet Decomposition Mathematcal Methods for Informaton Scence and Economcs Novel Pattern-based Fngerprnt Recognton Technque Usng D Wavelet Decomposton TUDOR BARBU Insttute of Computer Scence of the Romanan Academy T. Codrescu,,

More information

Cross-lingual Pseudo Relevance Feedback Based on Weak Relevant Topic Alignment

Cross-lingual Pseudo Relevance Feedback Based on Weak Relevant Topic Alignment Cross-lngual Pseudo Relevance Feedback Based on Weak Relevant opc Algnment WANG Xu-wen Insttute of Medcal Informaton & Lbrary, Chnese Academy of Medcal Scences, Beng 100020 wang.xuwen@mcams.ac.cn ZHANG

More information

Real-Time View Recognition and Event Detection for Sports Video

Real-Time View Recognition and Event Detection for Sports Video Real-Tme Vew Recognton and Event Detecton for Sports Vdeo Authors: D Zhong and Shh-Fu Chang {dzhong, sfchang@ee.columba.edu} Department of Electrcal Engneerng, Columba Unversty For specal ssue on Multmeda

More information

Problem Set 3 Solutions

Problem Set 3 Solutions Introducton to Algorthms October 4, 2002 Massachusetts Insttute of Technology 6046J/18410J Professors Erk Demane and Shaf Goldwasser Handout 14 Problem Set 3 Solutons (Exercses were not to be turned n,

More information

Deep Classification in Large-scale Text Hierarchies

Deep Classification in Large-scale Text Hierarchies Deep Classfcaton n Large-scale Text Herarches Gu-Rong Xue Dkan Xng Qang Yang 2 Yong Yu Dept. of Computer Scence and Engneerng Shangha Jao-Tong Unversty {grxue, dkxng, yyu}@apex.sjtu.edu.cn 2 Hong Kong

More information

Available online at Available online at Advanced in Control Engineering and Information Science

Available online at   Available online at   Advanced in Control Engineering and Information Science Avalable onlne at wwwscencedrectcom Avalable onlne at wwwscencedrectcom Proceda Proceda Engneerng Engneerng 00 (2011) 15000 000 (2011) 1642 1646 Proceda Engneerng wwwelsevercom/locate/proceda Advanced

More information

Linear Cross-Modal Hashing for Efficient Multimedia Search

Linear Cross-Modal Hashing for Efficient Multimedia Search Lnear Cross-Modal Hashng for Effcent Multmeda Search Xaofeng Zhu Z Huang Heng Tao Shen Xn Zhao College of CSIT, Guangx Normal Unversty, Guangx, 544,P.R.Chna School of ITEE, The Unversty of Queensland,

More information

Audio Content Classification Method Research Based on Two-step Strategy

Audio 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 information

A Novel Term_Class Relevance Measure for Text Categorization

A Novel Term_Class Relevance Measure for Text Categorization A Novel Term_Class Relevance Measure for Text Categorzaton D S Guru, Mahamad Suhl Department of Studes n Computer Scence, Unversty of Mysore, Mysore, Inda Abstract: In ths paper, we ntroduce a new measure

More information

Discriminative Dictionary Learning with Pairwise Constraints

Discriminative Dictionary Learning with Pairwise Constraints Dscrmnatve Dctonary Learnng wth Parwse Constrants Humn Guo Zhuoln Jang LARRY S. DAVIS UNIVERSITY OF MARYLAND Nov. 6 th, Outlne Introducton/motvaton Dctonary Learnng Dscrmnatve Dctonary Learnng wth Parwse

More information

Invariant Shape Object Recognition Using B-Spline, Cardinal Spline, and Genetic Algorithm

Invariant Shape Object Recognition Using B-Spline, Cardinal Spline, and Genetic Algorithm Proceedngs of the 5th WSEAS Int. Conf. on Sgnal Processng, Robotcs and Automaton, Madrd, Span, February 5-7, 6 (pp4-45) Invarant Shape Obect Recognton Usng B-Splne, Cardnal Splne, and Genetc Algorthm PISIT

More information

Query classification using topic models and support vector machine

Query classification using topic models and support vector machine Query classfcaton usng topc models and support vector machne Deu-Thu Le Unversty of Trento, Italy deuthu.le@ds.untn.t Raffaella Bernard Unversty of Trento, Italy bernard@ds.untn.t Abstract Ths paper descrbes

More information

Experiments in Text Categorization Using Term Selection by Distance to Transition Point

Experiments in Text Categorization Using Term Selection by Distance to Transition Point Experments n Text Categorzaton Usng Term Selecton by Dstance to Transton Pont Edgar Moyotl-Hernández, Héctor Jménez-Salazar Facultad de Cencas de la Computacón, B. Unversdad Autónoma de Puebla, 14 Sur

More information

Optimizing Document Scoring for Query Retrieval

Optimizing Document Scoring for Query Retrieval Optmzng Document Scorng for Query Retreval Brent Ellwen baellwe@cs.stanford.edu Abstract The goal of ths project was to automate the process of tunng a document query engne. Specfcally, I used machne learnng

More information

Collaboratively Regularized Nearest Points for Set Based Recognition

Collaboratively 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 information

A Deflected Grid-based Algorithm for Clustering Analysis

A Deflected Grid-based Algorithm for Clustering Analysis A Deflected Grd-based Algorthm for Clusterng Analyss NANCY P. LIN, CHUNG-I CHANG, HAO-EN CHUEH, HUNG-JEN CHEN, WEI-HUA HAO Department of Computer Scence and Informaton Engneerng Tamkang Unversty 5 Yng-chuan

More information

Exploring Image, Text and Geographic Evidences in ImageCLEF 2007

Exploring Image, Text and Geographic Evidences in ImageCLEF 2007 Explorng Image, Text and Geographc Evdences n ImageCLEF 2007 João Magalhães 1, Smon Overell 1, Stefan Rüger 1,2 1 Department of Computng Imperal College London South Kensngton Campus London SW7 2AZ, UK

More information

R s s f. m y s. SPH3UW Unit 7.3 Spherical Concave Mirrors Page 1 of 12. Notes

R s s f. m y s. SPH3UW Unit 7.3 Spherical Concave Mirrors Page 1 of 12. Notes SPH3UW Unt 7.3 Sphercal Concave Mrrors Page 1 of 1 Notes Physcs Tool box Concave Mrror If the reflectng surface takes place on the nner surface of the sphercal shape so that the centre of the mrror bulges

More information

Object-driven content-based image retrieval

Object-driven content-based image retrieval th Int. Worshop on Systems Sgnals & Image Processng -4 September 005 Chalda Greece 89 Obect-drven content-based mage retreval Ioanns Pratas* Baslos Gatos and Stavros Perantons Computatonal Intellgence

More information

Compiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz

Compiler 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 information

Combining Multiple Resources, Evidence and Criteria for Genomic Information Retrieval

Combining Multiple Resources, Evidence and Criteria for Genomic Information Retrieval Combnng Multple Resources, Evdence and Crtera for Genomc Informaton Retreval Luo S 1, Je Lu 2 and Jame Callan 2 1 Department of Computer Scence, Purdue Unversty, West Lafayette, IN 47907, USA ls@cs.purdue.edu

More information

Manifold-Ranking Based Keyword Propagation for Image Retrieval *

Manifold-Ranking Based Keyword Propagation for Image Retrieval * Manfold-Rankng Based Keyword Propagaton for Image Retreval * Hanghang Tong,, Jngru He,, Mngjng L 2, We-Yng Ma 2, Hong-Jang Zhang 2 and Changshu Zhang 3,3 Department of Automaton, Tsnghua Unversty, Bejng

More information

Information Retrieval

Information Retrieval Anmol Bhasn abhasn[at]cedar.buffalo.edu Moht Devnan mdevnan[at]cse.buffalo.edu Sprng 2005 #$ "% &'" (! Informaton Retreval )" " * + %, ##$ + *--. / "#,0, #'",,,#$ ", # " /,,#,0 1"%,2 '",, Documents are

More information

Proper Choice of Data Used for the Estimation of Datum Transformation Parameters

Proper 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 information

Real-time Motion Capture System Using One Video Camera Based on Color and Edge Distribution

Real-time Motion Capture System Using One Video Camera Based on Color and Edge Distribution Real-tme Moton Capture System Usng One Vdeo Camera Based on Color and Edge Dstrbuton YOSHIAKI AKAZAWA, YOSHIHIRO OKADA, AND KOICHI NIIJIMA Graduate School of Informaton Scence and Electrcal Engneerng,

More information

Research Article A High-Order CFS Algorithm for Clustering Big Data

Research Article A High-Order CFS Algorithm for Clustering Big Data Moble Informaton Systems Volume 26, Artcle ID 435627, 8 pages http://dx.do.org/.55/26/435627 Research Artcle A Hgh-Order Algorthm for Clusterng Bg Data Fanyu Bu,,2 Zhku Chen, Peng L, Tong Tang, 3 andyngzhang

More information

Fast Feature Value Searching for Face Detection

Fast Feature Value Searching for Face Detection Vol., No. 2 Computer and Informaton Scence Fast Feature Value Searchng for Face Detecton Yunyang Yan Department of Computer Engneerng Huayn Insttute of Technology Hua an 22300, Chna E-mal: areyyyke@63.com

More information

K-means and Hierarchical Clustering

K-means and Hierarchical Clustering Note to other teachers and users of these sldes. Andrew would be delghted f you found ths source materal useful n gvng your own lectures. Feel free to use these sldes verbatm, or to modfy them to ft your

More information

Assignment # 2. Farrukh Jabeen Algorithms 510 Assignment #2 Due Date: June 15, 2009.

Assignment # 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 information

Text Similarity Computing Based on LDA Topic Model and Word Co-occurrence

Text Similarity Computing Based on LDA Topic Model and Word Co-occurrence 2nd Internatonal Conference on Software Engneerng, Knowledge Engneerng and Informaton Engneerng (SEKEIE 204) Text Smlarty Computng Based on LDA Topc Model and Word Co-occurrence Mngla Shao School of Computer,

More information

Robust Mean Shift Tracking with Corrected Background-Weighted Histogram

Robust Mean Shift Tracking with Corrected Background-Weighted Histogram Robust Mean Shft Trackng wth Corrected Background-Weghted Hstogram Jfeng Nng, Le Zhang, Davd Zhang and Chengke Wu Abstract: The background-weghted hstogram (BWH) algorthm proposed n [] attempts to reduce

More information