Framework for Effective ANN and SVM Based Learning for Fast Multimedia Content-Based Retrieval

Ankush Mittal *
*Department of Electronics and Computer Engineering, Indian Institute of Technology, Roorkee, India.
Periodicity:October - December'2007
DOI : https://doi.org/10.26634/jse.2.2.541

Abstract

Recently, strategies involving learning a supervised model are emerging in the field of multimedia content-based retrieval. When there are clearly identified categories, as well as, large domain-representative training data, learning can be effectively employed to construct a model of the domain.

In this paper, an adequately domain-independent approach is presented where local features can characterize multimedia data using Artificial Neural Networks (ANN) and Support Vector Machines (SVM). The classification in content-based retrieval requires non-linear mapping of feature space. This can normally be accomplished by ANN and SVM. However, they inherently lack the capability to deal with meaningful feature evaluation and large dimensional feature space in the sense that they are inaccurate and slow. We overcome these defects by employing discrete Bayesian error based meaningful feature selection. The experiments on database consisting of real video sequences show that the speed and accuracy of SVM can be improved substantially using this technique, while execution time can be substantially reduced for ANN. The comparison also shows that improved SVM turns out to be a better choice than ANN. Finally, it is shown that generalization in learning is not affected by reducing the dimension of the feature space by the proposed method.

Keywords

Multimedia Indexing, Dimensionality Reduction, Neural Networks, Support Vector Machines, Fast Training, Relevant Features Evaluation

How to Cite this Article?

Ankush Mittal (2007). Framework for Effective ANN and SVM Based Learning for Fast Multimedia Content-Based Retrieval. i-manager’s Journal on Software Engineering, 2(2), 14-21. https://doi.org/10.26634/jse.2.2.541

References

[I ]. E. C. Yiu, Image classification using color cues and texture orientation, Master's thesls, Department of EECS, MIT, 1996.
[2]. M. Szummer and R. W Picard, Indoor-outdoor image classification. Int. Workshop on Content-based access of image and video databases, 1998,
[3]. M. M. Gorkhani and R. W. Picard, Texture orientation for sorting photos. /nter- nationa/ conference on Pattern recognltion, pages 459 -464, 1994,
[4]. A. B. Torralba and A. Oliva, Semantic organization of scenes using discriminant structural templates. Internatlona/ Conference on computer vlslon, pages 1253 -1258, vol, 2, 1999.
[5]. A. Vailaya and A. K. Jain, Reject option for VQ-based bayesian classification. Internationa/ Conference on pattern recognitlon, pages 48 - 5 I , 2000,
[6]. D. Forsyfh, J. Malik, M. Fleck, H. Greenspan, T. Leung, S. Belongie, C. Carson, and C. Bregler, Finding pictures of objects in large collections of images. Int. Workshop on Object Recognltlon for Computer Vlslon, pages 355 - 360, 1996 ,
[7]. s. Haykin, Neural network: A comprehensive foundation~ pages I 78 - 210, 2nd Edition, I 999~
[8]. S. Raudys, How good are support vector machines? Neural Networks, vo/. 13, pages I 7 - I 9, 2000,
[9]. R. R Lippman, A critical overview of neural network pattern classifiers. /EEE workshop on Neural Networks for s/gnal processlng, pages 266 275, 1991 ,
[l0]. Y.-H. Pao, in Adaptlve pattern recognltlon and neural networks, Addison-Wesley, 1989.
[I I ]. V. Vapnik, inThe nature of stat/stlcal leamlng theory, springer-verlag, I 995.
[I 2]. R Clarkson and R J. Moreno, On the use of support vector machines for phonetic classification. /EEE /nternatlonal Conference on Acoustlcs, Speech, and Signal Processlng, pages 585 -8, 1999
[I 3]. R. E. Karlson, D.J. Gorsich, and G . R. Gorhart, Target class! -cation via support vector machines. Opt/ca/ Englneerlng, pages 704 - I I , 2000, vol, 39 .
[I 4]. C. J. C. Surges, A tutorial on support vector machines for pattern recognition, Data m/nlng and knowledge dlscovery, pages 955 - 974, vol. 2 I 998.
[I 5]. W A. Khatib, Y F. Day, A. Ghafoor, and R 8. Serra, Semantic modeling and knowledge representation in multimedia databases, /EEE transact/ons on know/- edge and data englneerlng, pages 64- 80, 1999, vol, 1 I .
[I 6]. D. W. Aha and R. L. 8ankort, Feature selection for case-based classification of cloud types: An empirical comparison.AAA/, pages 106 - I I 2, 1994
[I 7]. H. Liu and R. Sotiono, Dimensionality reduction via discretization~ Know/edge based systems, pages 67- 72, no,9, 1 996.
[I 8]. R J. Devijvor and J. Kittler, Pattern recognition: A statistical approach, Prent/ce Ha//, 1982,
[I 9]. N~ Wyse, R. Dubos, and A. K. Jain, A critical evaluation of intrinsic dimensionality reduction algorithms, Pattern Recogn/tlon In practlse. Morgan Kaufmann., Pages 415 -425, 1980,
[20]. K. Fununaga, Introduction to statistical pattern recognition.Academ/cpress, 1990.
[21]. M. Modrzojowski, Selection using rough sets theory, European conference on Ma- chlne learnlng, pages 213 - 226, I 993.
[22]. H. Almuallim and T. G . Diottorich, Learning boolean concepts in the presence of many irrelevant features, Art/f/c/a/ /ntelllgence, pages 279 - 305, vol, 69, I 994.
[23]. M. Pazzani, C. Merz, K. Ali, and T. Hume, Reducing misclassification costs. /nternatlonal Conference on MachlneLeamlng, pages 217- 225, 1994,
[24]. M. Pazzani, An interatfve improvement approach for the discretization of numeric attributes in bayesian classifiers, Internat/onal Conference on Knowledge D/scovery and DataMlnlng(KDD}, pages 228 - 233, 1 995.
[25]. F. Garbor and A. Djouadi, Bounds on the bayes classification error based on pairwise risk functions~ /EEE transactlons on PAM/, vo/. 10, pages 281- 288, 1988.
[26]. K. Tumor and J. Ghosh, Estimating the bayes error rate through classifier combin- ing~ /nternat/ona/ Conference on Pattern recognltlon, pages 695-699, 1996
[27]. A. F. Kohn, L. G. Nakano, and M. O. Silva, A class discrimability measure based on feature space partitioning. Pattern recognltlon, pages 873 - 887, 1996 . 13
[28]. A. Mittal and L. F. Choong, Addressing the problems of bayesian network classifi- cation of video using high dimensional features, /EEE Transactlons on Knowledge and DataEnglneerlng, pages 230 - 244, 2004.
[29]. C. C. Chang and C.-J. Lin, LIBSVM: introduction and benchmarks, In Tech. Report, CS Deptt., NTU, Taiwan. http..Ilwww csle.ntu.edu.twl~cjllnlllbsvml, 2004,
If you have access to this article please login to view the article or kindly login to purchase the article

Purchase Instant Access

Single Article

North Americas,UK,
Middle East,Europe
India Rest of world
USD EUR INR USD-ROW
Pdf 35 35 200 20
Online 35 35 200 15
Pdf & Online 35 35 400 25

Options for accessing this content:
  • If you would like institutional access to this content, please recommend the title to your librarian.
    Library Recommendation Form
  • If you already have i-manager's user account: Login above and proceed to purchase the article.
  • New Users: Please register, then proceed to purchase the article.