LNCS 8321 Prosenjit Gupta Christos Zaroliagis (Eds.) Applied Algorithms First International Conference, ICAA 2014 Kolkata, India, January 2014 Proceedings 123 CuuDuongThanCong.com Lecture Notes in Computer Science Commenced Publication in 1973 Founding and Former Series Editors: Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen Editorial Board David Hutchison Lancaster University, UK Takeo Kanade Carnegie Mellon University, Pittsburgh, PA, USA Josef Kittler University of Surrey, Guildford, UK Jon M Kleinberg Cornell University, Ithaca, NY, USA Alfred Kobsa University of California, Irvine, CA, USA Friedemann Mattern ETH Zurich, Switzerland John C Mitchell Stanford University, CA, USA Moni Naor Weizmann Institute of Science, Rehovot, Israel Oscar Nierstrasz University of Bern, Switzerland C Pandu Rangan Indian Institute of Technology, Madras, India Bernhard Steffen TU Dortmund University, Germany Madhu Sudan Microsoft Research, Cambridge, MA, USA Demetri Terzopoulos University of California, Los Angeles, CA, USA Doug Tygar University of California, Berkeley, CA, USA Gerhard Weikum Max Planck Institute for Informatics, Saarbruecken, Germany CuuDuongThanCong.com 8321 Prosenjit Gupta Christos Zaroliagis (Eds.) Applied Algorithms First International Conference, ICAA 2014 Kolkata, India, January 13-15, 2014 Proceedings 13 CuuDuongThanCong.com Volume Editors Prosenjit Gupta Heritage Institute of Technology Computer Science and Engineering Chowbaga Road Anandapur, Kolkata 700107, India E-mail: prosenjit.gupta@heritageit.edu Christos Zaroliagis University of Patras Department of Computer Engineering and Informatics 26500 Patras, Greece E-mail: zaro@ceid.upatras.gr ISSN 0302-9743 e-ISSN 1611-3349 e-ISBN 978-3-319-04126-1 ISBN 978-3-319-04125-4 DOI 10.1007/978-3-319-04126-1 Springer Cham Heidelberg New York Dordrecht London Library of Congress Control Number: 2013956128 CR Subject Classification (1998): F.2, I.2, H.3-4, F.1, C.2, D.2, D.4.6 LNCS Sublibrary: SL – Programming and Software Engineering © Springer International Publishing Switzerland 2014 This work is subject to copyright All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s location, in its current version, and permission for use must always be obtained from Springer Permissions for use may be obtained through RightsLink at the Copyright Clearance Center Violations are liable to prosecution under the respective Copyright Law The use of general descriptive names, registered names, trademarks, service marks, etc in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use While the advice and information in this book are believed to be true and accurate at the date of publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made The publisher makes no warranty, express or implied, with respect to the material contained herein Typesetting: Camera-ready by author, data conversion by Scientific Publishing Services, Chennai, India Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com) CuuDuongThanCong.com Preface This volume contains papers accepted for presentation at the International Conference on Applied Algorithms (ICAA 2014) held at the Heritage Institute of Technology, Kolkata, India, during January 13–15, 2014, together with the extended or short abstracts of invited lectures given by Susanne Albers (Technical University of Munich, Germany), Bhargab Bhattacharya (Indian Statistical Institute, Kolkata, India), Gautam Das (University of Texas at Arlington, USA), Dimitrios Gunopulos (University of Athens, Greece), Rina Panigrahy (Microsoft Research, Mountain View, USA), Assaf Schuster (Technion, Israel Institute of Technology, Haifa, Israel), and Christos Zaroliagis (CTI & University of Patras, Greece) ICAA is a new conference series with a mission to provide a quality forum for researchers working in applied algorithms Papers presenting original contributions related to the design, analysis, implementation, and experimental evaluation of efficient algorithms and data structures for problems with relevant real-world applications were sought, ideally bridging the gap between academia and industry Papers were solicited describing original research in a variety of areas including (but not limited to): – – – – – – – – – – – – – – – – – – – – – – – Algorithmic Microfluidics Algorithms for VLSI CAD Analysis of Algorithms Approximation Algorithms Big Data Algorithms Cloud Computing Computational Advertising Computational Biology Computational Geometry Computational Services Science Computational Transportation Science Cryptography and Security Databases Data Mining Data Structures Distributed Algorithms Energy Efficient Algorithms External Memory Algorithms Graph Algorithms Graph Drawing Hardware Accelerated Algorithms Heuristic Search Image Processing CuuDuongThanCong.com VI – – – – – – – – – – – – – – Preface Information Retrieval Location Based Services Machine Learning Parallel Algorithms Pattern Recognition Railway Optimization Randomized Algorithms Recommender Systems Robotics Spatial Informatics Social Network Analysis Web Intelligence Web Mining Web Searching In response to the call-for-papers, 122 submissions from countries were received The Program Committee (comprising of 31 members from countries) selected 21 papers for presentation The criteria for selection were perceived originality, quality, and relevance to the subject area of the conference Considerable effort was devoted to the evaluation of the submissions and to providing authors with helpful feedback Towards this end, the Program Committee was assisted by 45 external reviewers We thank all those who submitted papers for consideration, as well as the Program Committee members and external reviewers for their invaluable contribution We thank the management of the Heritage Institute of Technology and the Kalyan Bharathi Trust and the entire Organizing Committee for the excellent arrangements leading up to and during the entire conference Finally, we gratefully acknowledge the generous financial support received from the TEQIP grant which made this conference a possibility January 2014 CuuDuongThanCong.com Prosenjit Gupta Christos Zaroliagis Organizing Committee Heritage Institute of Technology, Kolkata, India Advisors H.K Chaudhary, Chairman P.K Agarwal, CEO Probir Roy, Executive Director B.B Paira, Advisor D.C Ray, TEQIP-II Coordinator S.N Biswas, Deputy Director Sukumar Ghosh, Professor Kalyan Bharti Trust Kalyan Bharti Trust Kalyan Bharti Trust Higher Education, Kalyan Bharti Trust Heritage Institute of Technology Heritage Institute of Technology University of Iowa, USA General Chairs Pranay Chaudhuri, Principal Amitava Bagchi, Professor Organizing Chairs Kalarab Ray Tapan Chakraborty Subhashis Majumder Subcommittee Chairs Satarupa Bagchi Biswas Dinabandhu Bhandari Arindam Chatterjee Poulami Das Aniruddha Dasgupta Hemanta K De Anindita Kundu Anindya Jyoti Pal Shilpi Saha Sujay Saha Somenath Sengupta Arvind Srivastava CuuDuongThanCong.com Heritage Institute of Technology Department of Computer Science and Engineering VIII Organizing Committee Local Hospitality All faculty members and technical assistants of The Department of Computer Science and Engineering and The Department of Information Technology, Heritage Institute of Technology, Kolkata, India Program Committee Bogdan Arsintescu Mukul Bansal Kostas Berberidis Bhargab Bhattacharya Danny Chen Jinjun Chen Anandaswarup Das Gautam Das Nabanita Das Aniruddha Dasgupta Bhaskar Dasgupta Dimitrios Gunopulos Prosenjit Gupta Ravi Janardan C.V Jawahar Matthew Katz Spyros Kontogiannis Vamsi Kundeti Subhashis Majumder Neeraj Mittal Srihari Nelakuditi Nikos Nikolaidis Rina Panigrahy Sudeshna Sarkar Anup Sen Michiel Smid Jack Snoeyink Kannan Srinathan Kostas Tsichlas Dorothea Wagner Christos Zaroliagis CuuDuongThanCong.com Google, Mountain View, USA MIT University of Patras, Greece Indian Statistical Institute, India University of Notre Dame, France University of Technology, Australia IBM Research University of Texas at Arlington, USA Indian Statistical Institute, India Heritage Institute of Technology, India University of Illinois at Chicago, USA National and Kapodistrian University of Athens, Greece Heritage Institute of Technology, India University of Minnesota, USA IIIT Hyderabad, India Ben-Gurion University of the Negev, Israel University of Ioannina Intel Heritage Institute of Technology, India University of Texas at Dallas, USA University of South Carolina, USA Aristotle University of Thessaloniki, Greece Microsoft Research IIT Kharagpur, India Indian Institute of Management, India Carleton University, Canada University of North Carolina at Chapel Hill, USA IIIT Hyderabad, India Aristotle University of Thessaloniki, Greece Karlsruhe Institute of Technology, Germany University of Patras, Greece Organizing Committee Additional Reviewers Abu-Affash, A Karim Aschner, Rom Bagchi, Amitava Banerjee, Ansuman Banerjee, Sabyasachee Barash, Danny Basu Chowdhuri, Partha Basuchowdhuri, Partha Baum, Moritz Blă asius, Thomas Carmi, Paz Chatterjee, Arindam Chen, Jianxu Christodoulakis, Manolis Das, Poulami Fotakis, Dimitris Freeman, Clinton Fuchs, Fabian Gallopoulos, Efstratios Gounaris, Anastasios Iakovidou, Nantia Kappes, Andrea Katsaros, Dimitrios CuuDuongThanCong.com Kondapally, Ranganath Konstantinou, Elisavet Kosmatopoulos, Andreas Kundu, Malay Kumar Majumder, Prasenjit Mchedlidze, Tamara Mu, Jian Nandy, Subhas Nikoletseas, Sotiris Nă ollenburg, Martin Papadopoulos, Apostolos Prutkin, Roman Ramachandran, Arunmoezhi Rapti, Angeliki Saha, Sujay Segal, Michael Sioutas, Spyros Sur-Kolay, Susmita Tefas, Anastasios Tong, Yan Wang, Jiazhuo Xue, Yuan Zhang, Yizhe IX Abstracts of Invited Talks CuuDuongThanCong.com 264 A Nag et al Fig The cover image Fig The stego image Table Capacity and PSNR for different Images Images Size of cover image Capacity of embedding Name Pixel Pixel Lena 256 128 Baboon 256 128 Airplane 256 128 Boat 256 128 PSNR dB 30.48 30.28 30.91 30.36 will be an encoded message which will be useless to them The attacker needs to find out the public key and Huffman Table in order to decrypt that message and get the actual content Conclusion In this paper, we have proposed a novel steganographic method based on Huffman code and location based LSB substitution Secret data are encoded into each pixel of cover image by 4-bit LSB modification method We not embed the actual data instead we change the bit values of certain position in LSB of the cover image Along with this encoding, we also perform encoding of the secret message using Huffman code which makes it more secure than existing steganographic techniques Experiments show that the stego-image of our method are almost identical to the cover image The stego image generated by our method has got just one in LSB of the stego image which can be a point of attack by steganalyzers The authors are currently engaged into finding ways to mitigate this limitation and make a more robust signature free stego image CuuDuongThanCong.com A Huffman Code Based Image Steganography Technique 265 References Cheddad, A., Condell, J., Curran, K., McKevitt, P.: Digital image steganography: Survey and analysis of current methods Signal Processing 90, 727–752 (2010) Nag, A., Biswas, S., Sarkar, D., Sarkar, P.P.: A novel technique for image steganography based on block-DCT and Huffman encoding International Journal of Computer Science and Information Technology 2(3), 103–111 (2010) Nag, A., Biswas, S., Sarkar, D., Sarkar, P.P.: A novel technique for image steganography based on DWT and Huffman encoding International Journal of Computer Science and Security 4(5), 561–570 (2010) Bender, D.W., Gruhl, N.M., Lu, A.: Techniques for data hiding IBM Systems Journal 35, 313–316 (1996) Wang, R.Z., Lin, C.F., Lin, J.C.: Image hiding by optimal LSB substitution and genetic algorithm Pattern Recognition 34(3), 671–683 (2001) Wu, H.C., Wu, N.I., Tsai, C.S., Hwang, M.S.: Image steganographic scheme based on pixel-value differencing and lsb replacement methods Images Signal Processing 152(5), 611–615 (2005) Park, Y.-R., Kang, H.-H., Shin, S.-U., Kwon, K.-R.: A steganographic scheme in digital images using information of neighboring pixels In: Wang, L., Chen, K., Ong, Y.S (eds.) ICNC 2005 LNCS, vol 3612, pp 962–967 Springer, Heidelberg (2005) Wu, D.C., Tsai, W.H.: A steganographic method for images by pixel-value differencing Pattern Recognition Letters 24(9-10), 1613–1626 (2003) Chang, C.C., Tseng, H.W.: A steganographic method for digital images using side match Pattern Recognition Letters 25(12), 1431–1437 (2004) 10 Wang, C.-M., Wu, N.-I., Tsai, C.-S., Hwang, M.-S.: A high quality steganographic method with pixel-value differencing and modulus function Journal of System Software 81, 150–158 (2008) 11 Yang, C.-H., Weng, C.-Y., Wang, S.-J., Sun, H.-M.: Adaptive data hiding in edge areas of images with spatial LSB domain systems IEEE Transactions on Information Forensics and Security 3(3), 488–497 (2008) 12 Liao, X., Wen, Q.-Y., Zhang, J.: A steganographic method for digital images with four-pixel differencing and modified lsb substitution Journal Visual Communication and Image Representation 22, 1–8 (2011) 13 McIntyre, D.R., Pechura, M.A.: Data compression using static Huffman codedecode tables Communications of ACM 28, 612–616 (1985) 14 Jeong, J.C., Jo, J.M.: Adaptive huffman coding of 2-d dct coefficients for image sequence compression SP:IC 7(1), 1–11 (1995) 15 Rivest, R., Shamir, A., Adleman, L.: A method for obtaining digital signatures and public-key crypto-systems Communications of the ACM 21(2), 120–126 (1978) 16 Stallings, W.: Cryptography and Network Security: Principles and Practices, 4th edn Pearson Education Pvt Ltd., India (2004) CuuDuongThanCong.com Expression-Invariant 3D Face Recognition Using K-SVD Method Somsukla Maiti1,2 , Dhiraj Sangwan2 , and Jagdish Lal Raheja1,2 Academy of Scientific and Innovative Research CSIR-Central Electronics Engineering Research Institute, Pilani, Rajasthan, India somsuklamaiti@gmail.com, {dhiraj,jagdish}@ceeri.ernet.in Abstract This paper proposes a method to perform expression invariant face recognition using dictionary learning approach The proposed method performs the operation in the following stages: the T-region extraction from the face to get the facial region having minimum variation with expression, determination of the wavelet coefficients of the extracted region, dictionary learning using K-SVD and matching The experiment has been performed on a database that contains 40 persons with expressions each under different illumination conditions The recognition performed has shown a good accuracy rate as compared to the mostly used PCA-SVM approach Our system uses label-consistent K-SVD algorithm for dictionary learning to learn a set of dictionaries that represents 3D information of the face This method fulfills the purpose of sparse coding and classification Keywords: Label-Consistent KSVD, Sparse Coding, Expression Invariant, T-Region, Dictionary Learning Introduction Most of the recent security issues have brought up the attention of mankind about the serious drawbacks in most of the sophisticated security systems To solve these issues, the security systems are heading towards the use of biometric technologies for verification and identification of individuals Biometrics refers to the identification of humans by their physical traits The biometric system, in general, processes the raw data captured from camera, scanner, RFID Tag etc Certain features are then extracted from the data followed by the extraction of templates which are easier to process and store, but carries most of the important information needed about the person There are several biometric traits that are used commonly for the recognition of individuals Face recognition is one of the most acceptable and popular method, as it is a good tradeoff between reliability and social acceptance due to its non-intrusive nature; and balances security and privacy at the same time [1] However, the face recognition suffers from some significant challenges due to variations in illumination, viewing angle, facial expressions, occlusion, and changes over time, etc P Gupta and C Zaroliagis (Eds.): ICAA 2014, LNCS 8321, pp 266–276, 2014 c Springer International Publishing Switzerland 2014 CuuDuongThanCong.com Expression-Invariant 3D Face Recognition Using K-SVD Method 267 In the past two decades, most of the works were focused on 2D facial images But the current 2D face recognition systems are greatly affected by differences in pose, illumination, expressions, and other characteristics that can vary between the captures of a human face This issue becomes more significant when the subject has incentives of not to be recognized There are very few 2D algorithms which handle variations in pose and illumination There exists Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA) and Independent Component Analysis (ICA) based algorithms for 2D face recognition All these statistical methods not consider the effects of facial expressions and very large variations in pose To achieve significantly higher accuracy 3D face recognition methods have been adopted The 3D facial data can provide a promising way to understand the feature of the human face in 3D space and has potential possibility to improve the performance of the system There are some distinct advantages in using 3D information: sufficient geometrical information, invariance of measured features relative to transformation and capture process by laser scanners being immune to illumination variation Several studies have been performed in 3D face recognition Early face recognition algorithms advocated the invariant approach by finding a set of fiducial points such as eyes, nose, mouth, etc and comparing their geometric relations (feature-based recognition) or comparing the face with a whole facial template (template-based recognition) Xu et al suggested an automatic face recognition method combining the global geometric features with the local shape variation information of the face [2] The method represents the face using a scattered 3D point cloud with a regular mesh by using the hierarchical mesh fitting For the purpose of recognition, the mesh data and the shape information had been passed to dimensionality reduction chamber using principal component analysis (PCA) [3, 4] to obtain a lower-dimensional vector followed by nearest neighbor classifier (NN) for classification The method does not provide a good degree of accuracy in recognition and the parameters varies drastically with the different sets of faces Another approach of 3D face recognition was suggested by Bronstein et al that concentrated on the computation of the geodesic distances over the facial surfaces, given only the metric tensor of the surface [5] To determine the face features the method of Fast Marching (FMM) had been used The method though provides a good result, but leads to lot of computations to be performed in finding the geodesic distances over individual face surfaces The expression and pose invariant face recognition has captured the major atten-tion in the recent researches of face recognition This is to provide more robustness to the system and to make it view independent The 3D model-based pose invariant face recognition method estimated the pose from the face view available and then tried to adapt a 3D face model [6] Thus the frontal view face images were synthesized using the estimated 3D models and the discriminant features were further extracted from these synthesized frontal view images In this paper, the wavelets are used as the features and are classified using nearest CuuDuongThanCong.com 268 S Maiti, D Sangwan, and J.L Raheja feature space classifier The algorithm provides a robust recognition and can recognize faces under variable poses with good accuracy Berretti et al presented a novel approach in the recent time (2010) to 3D face matching that showed high effectiveness in distinguishing the facial differences be-tween distinct individuals from differences induced by non-neutral expressions within the same individual [7] The face is partitioned into iso-geodesic stripes that provide an approximate representation of the local morphology of faces that exhibits smooth variations for changes induced by facial expressions The approach takes into account the geometrical information of the 3D face and encodes the relevant information in the form of a graph 1.1 Proposed Method The method proposed here is based on the idea of dictionary learning using the Label-Consistent KSVD As the facial expressions have a large impact on the mouth region of the face so the T-region of the face has been extracted from the face The T-region is least affected and contains the most important information in the face T-region specifies the region starting from the top of the eyebrows to the nose end The wavelets of the T-region have been calculated to find the characteristics/features of the face The method allows memory efficient representation of the face with the maximum information content in it The process then uses label-consistent K-SVD algorithm for dictionary learning to learn a set of dictionaries to perform sparse cod-ing and classification of the 3D information of the faces from the database Label Consistent K-SVD The face data is obtained as high dimension large size matrix The underlying features that can be used to recognize the face are large in number and often contain multiple correlated versions of the same feature The relevant information about the faces is generally of much smaller dimensionality and so the first aim is to reduce the redundancy in data and get the relevant information Dictionary learning [8][9] is a method to determine the proper representation of data sets by means of reduced dimensionality subspaces, which are adaptive to both the characteristics of the input signals and the processing task at hand These representations are based on the principle that our observations can be described by a sparse subset of atoms taken from a redundant dictionary that features the main difference among the faces in the database Let us consider Y = yi , i = 1, 2, , N as a set of n dimensional N input signals yi ∈ RnXN Learning a reconstructive dictionary with K items for sparse representation of Y (where N >> K) can be accomplished by solving (1) < D, X >= arg ||Y − DX||2 subject to ∀i, ||xi ||0 = arg ||Y − DX||2 + α||Q − AX||2 D,A,X (2) where α is the parameter that controls the relative contribution between the reconstruction (the first term) and the label consistent regularization(the second term) The term ||Y − DX||2 is the reconstruction error and the term ||Q − AX||2 represents the discriminative sparse code error and the term Q is CuuDuongThanCong.com 270 S Maiti, D Sangwan, and J.L Raheja called the label consistent matrix that enforces dictionary items from the same label as the input signals to be used in the reconstruction of such signal It is defined as Q = q1 , q2 , , qN ∈ RN XK those are the discriminative sparse codes of input signals Y for classification To make the dictionary optimal for classification, the classification error is considered as a term in the objective function as mentioned earlier Thus the objective function for learning a dictionary D having both reconstructive and discriminative nature has been defined in (3) < D, W, A, X > = arg D,W,A,X 2 ||Y − DX||2 + α||Q − AX||2 + β||H − W X||2 subject to ∀i, ||xi ||0 = arg D ,W ,X 2 ||Y −D X ||2 + α||Q−X ||2 + β||H −W X ||2 subject to ∀i, ||xi ||0 = arg || ⎝ √ αQ ⎠ − ⎝ √ αA ⎠ X||2 D,W,A,X βH βW subject to ∀i, ||xi ||0