Projects

Project:	Project Definition:	Related files/links:
0) Activity detection in location-based datasets [Image to come soon.]	In this project we develop techniques to process large location-based datasets for the detection of various types of activities. We develop and apply coreset based techniques to summarize and speed up the datasets. More info to come soon here!
1) Visual object recognition in large datasets	If you ever used Google's image search engine by uploading an image, you may have noticed that, while it works quiet impressive, it is still far from being perfect (or near perfect). The fundamental problem in such image based search systems is defining and describing the most "similar image". A succesful image based search system would first analyze the contents of the given input image, and then lists the images including the most similar content among all the available images in a given database (such as Google's image database) My current research includes working on various steps of such an image retrieval system. These steps include image analysis, feature computation, feature selection and classification. I will include more details in the near future here.
2) Activity Detection in Scientific Visualization	In this project we look for a way to "extract" an "activity" in time varying scientific data sets. As an example, consider the videos shown on the left. The data set contains a Computational Fluid Dynamics (CFD) simulation formed of 100 time steps. The variable being visualized is the vorticity magnitude. The first video visualizes the evolution of all the extracted vortices over the time. Now in this data set, assume that you are not interested in seeing all the vortices, and instead, you are looking for only the vortices performing a specific activity. Consider that activity is the "merge-split" activity in which individual vortices merge first, and then this new "combined feature" splits again within the next 5 time steps. Searching manually for all the instances (or maybe even for a single instance) of this activity is a hard task even in this small data set. An automated process would extract all the instances of the activity easily. Therefore, in this project, we look for a way to describe, model an activity first, and then extract all the instances of such activity within the data set. The second video highlights the instances of the "merge-split" activity found in the data set using our technique. For more information on this project, check my Ph.D. dissertation or the TVCG paper: Activity Detection in Scientific Visualization .
3) Group Tracking in Scientific Visualization	Similar to the computer vision where the purpose is mainly analysing the video data of humans, there are groups in scientific data sets. A group is fundamentally a set of objects that are related "somehow". This relation can be only logical (such as a group of all green objects) or also physical (such as a flock of birds, a school of fish or a set of stars,i.e. a galaxy). The definition of a group changes from domain to domain, and this makes it harder to find a generic framework that could help scientists to define, extract and follow the evolution of groups in their data sets. Moreover, as the data dimensions increase, it becomes more and more apparent that smarter and more meaningful abstractions are necessary in large data visualization. In this work, to help with the above-mentioned problems, we propose using similarity functions to define a (physical) group in scientific visualization. Similarity functions can map the physical group definitions in scientific domains into the computational domain. Once a similarity function is set (defined), a clustering algorithm can be used to extract the groups in scientific simulations based on the similarity function. When groups are determined, it is necessary to track them to understand their evolution over time (in time varying data sets). Therefore in this study, we proposed a group tracking schema in which we track groups of objects (features) as well as the invidual objects. Group tracking allows us to define and detect new type of events such as cross-group, partial or full merge and partial or full split. The older versions of the feature (and group) tracking code are available on the old feature-tracking website (the link is on the right). The more recent (and fast working) version is available through Vizlab, Rutgers or through me.
4) Prostate Cancer Segmentation in MultiSpectral-MRI data	In this project we (along with our collaborators from University of Toronto) worked on using various data types that comes from different imaging modalities to detect prostate cancer and checked if combining the information from such different modalities would increase the accurracy of the segmentation of the prostate cancer in automated methods. Our results indicated that, indeed using more than one modality in MRI data, increases the accuracy of prostate cancer segmentation with automated tools. In this work, we showed that supervised techniques can work better than unsupervised techniques for prostate cancer segmentation. We discussed that using the traditional thresholding schema is not always the best solution for classifier based prostate cancer segmentation techniques. Therefore, we proposed various techniques for the training part to find "a better" threshold value that increases the segmentation efficiency for a given criteria such as the best accuracy, the best dice measure, the best sensitivity value, etc. In this work, we used Support Vector Machines, Relevance Vector Machines (Sparse Bayesian Learning) and Markov Random Fields to segment prostate cancer. We compared the results by computing the area under the curve (AUC) values for each classifier. In this work we demonstrated that using AUC values are more reliable when comparing the performances of different classifiers as opposed to just comparing the standard outputs since changing the threshold of a classifier is relatively a simple task.
5) "A Kernel study on Support Vector Machines"	In this work, we presented a way to increase the accuracy of the previously presented Chebshev kernels. In order to do that, we first presented a way to apply Chebyshev polynomials on vector inputs by introducing "Generalized Chebyshev Polynomials". And then, with examples, we demonstrated that how the proposed generalized Chebyshev polynomials (of both first and second kinds) can be used to create a set of kernel functions. Moreover, in this work we studied the SVM performance vs. kernel parameter and found out that for some kernels the performance can be changed drastically with the kernel parameter when compared to the other kernel functions. In this work, we also discussed some properties of generic kernel functions where we claimed that first applying the kernel functions on individual features (in a vector) and then multiplying the results to obtain the final result for a pair of given two input vectors, is not a good approach to use in SVM learning in general. As a summary in this work, we presented tools, techniques and examples to generate new kernel functions using Chebyshev polynomials. And discussed that SVM with Gaussian kernels can be considered as a clustering algorithm that focuses on the edges between two given classes. I have posted some MATLAB files related to the Chebyshev polynomials and Chebyshev kernels on the right. There are two .m files that does the same work as a Chebyshev kernel on the right (gencheb.m and chebkernel.m files). I wrote these files quickly for those who would like to study Chebyshev kernels. And they both are expected to provide the same result. However, please keep in mind that none is (computationally) the optimal code. And again, as described on my Pattern Recognition paper, please remember that, especially for the generalized Chebyshev kernel, you might need a normalization step prior to using the kernel functions. Each file has its own description within itself. Use them at your own risk, no warranty is provided! :)	gencheb.m chebkernel.m chevpoly.m chevpolyu.m

Publications

While my peer-reviewed journal & conference papers are listed below, you can always access the most up-to-date list on my Google Scholar page. Each individual link directs to the related Google scholar page (or to the publisher's related webpage) where you can get more information about the listed paper (such as its abstract or citations). Alternatively, you can also click on the [PDF] link next to the listed paper to download the paper.

Peer-Reviewed Journals: 1) K. G. Bemis, D. Silver, G. Xu, R. Light, D. Jackson, C. Jones, S. Ozer, L. Lui, The Path to COVIS: a review of acoustic imaging of hydrothermal flow regimes " Journal of Deep-Sea Research Part II, Vol. 121, pp.159,176 ,2015. 2) S. Ozer, D. Silver, K. Bemis, P. Martin, " Activity Detection in Scientific Visualization ", Visualization and Computer Graphics, IEEE Transactions on, vol.20, no.3, pp.377,390, March 2014, doi: 10.1109/TVCG.2013.117. [PDF] 3) S. Ozer, C.H. Chen, H.A. Cirpan, "A Set of New Chebyshev Kernel Functions for Support Vector Machine Pattern Classification", Journal of Pattern Recognition, 44 (7), 1435-1447, 2011. [PDF] 4) S. Ozer, D. L. Langer, X. Liu, M. A. Haider, T. H. van der Kwast, A. J. Evans, Y. Yang, M. N. Wernick, I. S. Yetik, "Supervised and Unsupervised Methods for Prostate Cancer Segmentation With Multispectral MRI ", Journal of Medical Physics, Vol. 37, Issue:4, April 2010. [PDF]
Peer-Reviewed Conferences, Workshops, Symposia: 13) S. Ozer, "Visual Object Recognition with Image Retrieval", Handbook of Pattern Recognition & Computer Vision, World Scientific, (invited book chapter), 2015. 12) R. Sarkar, S. Ozer, S. Acton, K. Skadron, "Image classification by multi-kernel dictionary learning”", IEEE Asilomar Conference on Signals, Systems and Computers, 2014. 11) S. Ozer, D. Silver, K. Bemis, P. Martin, "Activity Detection in Scientific Visualization (Invited Talk)" IEEE Visualization, 2014. 10) L. Liu, S. Ozer, K. Bemis, J. Takle, D. Silver, "An interactive method for activity detection visualization", Large Data Analysis and Visualization (LDAV), 2013 IEEE Symposium on, 2013 (Poster). [PDF] 9) K.G. Bemis, S. Ozer, G. Xu, P.A. Rona, D. Silver, "Event Detection for Hydrothermal Plumes: A case study at Grotto Vent", AGU Fall Meeting Abstracts 1, 05, 2012. 8) S. Ozer, J. Wei, D. Silver, K.-L. Ma, P. Martin, "Group Dynamics in Scientific Visualization", Large Data Analysis and Visualization (LDAV), 2012 IEEE Symposium on, 2012. [PDF] 7) S. Ozer, D. Silver, K. Bemis, P. Martin, J. Takle, "Activity Detection for Scientific Visualization", Large Data Analysis and Visualization (LDAV), 2011 IEEE Symposium on, 117-118, 2011 (Poster). [PDF] 6) S. Ozer, D. Silver, P. Martin, "Group Tracking in Scientific Visualization", IEEE Visualization, Visweek, Providence, RI, 2011 (poster). [PDF] 5) S. Ozer, M.A. Haider, D.L. Langer, T.H. van der Kwast, A.J. Evans, M.N. Wernick, J. Trachtenberg, I.S. Yetik, "Prostate Cancer Localization with Multispectral MRI Based on Relevance Vector Machine", IEEE International Symposium on Biomedical Imaging, ISBI09, Boston, MA, June28-July01, 2009. [PDF] 4) S. Ozer, C.H. Chen, I.S. Yetik, "Using K-NN SVMs for Performance Improvement and Comparison to K-highest Lagrange Multipliers Selection", joint IAPR International Workshops on Structural and Syntactic Pattern Recognition and Statistical Techniques in Pattern Recognition, SSPR 2010, Springer- Verlag, Lecture Notes in Computer Science, August 2010. [PDF] 3) S. Ozer, C. H. Chen, "Generalized Chebyshev Kernels for Support Vector Classification", ICPR 08, International Conference on Pattern Recognition, 7-11 December 2008, Tampa-FL, USA. [PDF] 2) S. Ozer, H. A. Cirpan, N. Kabaoglu, "Support Vector Regression for Surveillance Purposes", MRCS - International Workshop on Multimedia Content Representation, Classification and Security, Springer- Verlag, Lecture Notes in Computer Science, Pages:442-449, 11-13 September 2006, Istanbul, Turkey. 1) S. Ozer, H. A. Cirpan, N. Kabaoglu, "Support Vector Machines Based Target Tracking Techniques" (in Turkish), IEEE SIU2006, Sinyal Isleme ve Uygulamalari Kurultayi, 17-19 April 2006, Antalya, Turkiye.
Ph.D. Dissertation: "Activity Detection in Scientific Visualization", 2013: A Ph.D. dissertation in Electrical and Computer Engineering (Rutgers University, NJ) that studied the definition of an activity, its modelling and application issues in scientific (3D-time varying) simulations. M.Sc. Thesis: "On the Classification Performance of Support Vector Machines Using Chebyshev Kernel Functions":A Thesis in Electrical Engineering, S. Ozer, University of Massachusetts Dartmouth, 2007. (Upon request, I can send you a PDF version of my M.Sc. thesis). My "A Set of New Chebyshev Kernel Functions for Support Vector Machine Pattern Classification" journal paper reports most of the important parts of the thesis. Therefore I recommend reading the journal paper. (which can be obtained freely through many universities).
	Book: Sedat Ozer and Chi Hau Chen, "A kernel study on Support Vector Machines: Generalized Chebyshev Kernels", 2009. LAP Lambert Academic Publishing , ISBN-10: 3838305574 | ISBN-13: 978-3838305578. About the book: Mainly, this is the published version of my M.Sc. thesis from University of Massachusetts, Dartmouth. The book may have some minor grammatical mistakes in it, since I had to check the grammar and edit it within a very limited time. Besides such mistakes, if you have any thoughts, questions or feedback about the book, I would be more than happy to hear them. The book initially (and briefly) describes what machine learning is (specifically in the supervised-learning sense), and then goes into the theory of SVM, how the dual form of the SVM cost function is derived. Then it describes the previous work on Chebyshev kernels and then how to generalize them for vector inputs. Finally, it compares various kernel functions and their performances on various benchmark data sets including our proposed generalized Chebyshev kernel functions. My "A Set of New Chebyshev Kernel Functions for Support Vector Machine Pattern Classification" journal paper reports most important parts of the book. Therefore I recommend reading the journal paper. (which can be obtained freely through many universities).

Taken Courses

Taken courses: In the below table, I have listed "only" the important courses that I have taken along my B.Sc., M.Sc. and Ph.D. studies at various universities, and I believe these are the set of courses that can identify my background altogether. (Note: for each of the courses listed below, I have spent a considerable amount of my life where I had taken some courses more than once for a deeper understanding. :).
Signal Processing & Systems Related Courses:	Communication & RF (Radio-Frequency) Related Courses:	Image Processing and Pattern Recognition Related Courses:	Computer Systems & Software Engineering Related Courses:
Probability Theory and Random Processes Digital Speech Processing Digital Signal Processing Digital Signal Processing II Random Signals & Systems I Linear Systems Mathematics of Systems Analysis Signal Processing Apllications (in Matlab) Adaptive Filters Estimation Theory Statistical Signal Processing	Wireless Communication Systems Communication Engineering Fundamentals Mobile Communications Wireless Communication in Mobile Networks Electromagnetics Electromagnetic Field Theory Electromagnetic Wave Theory Transmission Lines Microwave Techniques Antennas Antennas and Microwave Lab.	Pattern Recognition Neural Networks Computer Vision & Image Processing Imaging Theory & Applications Digital Image Processing Fuzzy Logic and Fuzzy Systems Optimization Methods	Introduction to Parallel & Distributed Computing Software Engineering Software Engineering and Web Applications Algorithms and Data Structures Computer Networks Microprocessors and Computers System Design With Microcontrollers Computer Sytems Special Problems (focusing on Scientific Visualization)