Publications without Fulltext
Permanent URI for this collectionhttps://hdl.handle.net/20.500.14288/3
Browse
1703 results
Search Results
Publication Metadata only Comparison of convex combination and affine combination of adaptive filters(Ieee, 2009) Singer, Andrew C.; Department of Electrical and Electronics Engineering; Department of Electrical and Electronics Engineering; Kozat, Süleyman Serdar; Erdoğan, Alper Tunga; Faculty Member; Faculty Member; Department of Electrical and Electronics Engineering; College of Engineering; College of Engineering; 177972; 41624In the area of combination of adaptive filters, two main approaches, namely convex and affine combinations have been introduced. In this article, the relation between these two approaches is investigated. First, the problem of obtaining optimal convex combination coefficients is formulated as the projection of the optimal affine combination weights to the unit simplex in a weighted inner product space. Based on this formulation the closed form expressions for optimal combination weights and target MSE levels are obtained for two and three branch cases.Publication Metadata only Subspace-based techniques for retrieval of general 3D models(IEEE, 2009) Sankur, Bülent; Dutaǧac, Helin; Department of Computer Engineering; Yemez, Yücel; Faculty Member; Department of Computer Engineering; College of Engineering; 107907In this paper we investigate the potential of subspace techniques, such as, PCA, ICA and NMF in the case of indexing and retrieval of generic 3D models. We address the 3D shape alignment problem via continuous PCA and the exhaustive axis labeling and reflections. We find that the most propitious 3D distance transform leading to discriminative subspace features is the inverse distance transform. Our performance on the Princeton Shape Benchmark is on a par with the state-of-the-art methods. ©2009 IEEE.Publication Metadata only Performance measures for video object segmentation and tracking(IEEE-Inst Electrical Electronics Engineers Inc, 2004) Erdem, Çiğdem Eroğlu; Sankur, Bülent; Department of Electrical and Electronics Engineering; Tekalp, Ahmet Murat; Faculty Member; Department of Electrical and Electronics Engineering; College of Engineering; 26207We propose measures to evaluate quantitatively the performance of video object segmentation and tracking methods without ground-truth (GT) segmentation maps. The proposed measures are based on spatial differences of color and motion along the boundary of the estimated video object plane and temporal differences between the color histogram of the current object plane and its predecessors. They can be used to localize (spatially and/or temporally) regions where segmentation results are good or bad; and/or they can be combined to yield a single numerical measure to indicate the goodness of the boundary segmentation and tracking results over a sequence. The validity of the proposed performance measures without GT have been demonstrated by canonical correlation analysis with another set of measures with GT on a set of sequences (where GT information is available). Experimental results are presented to evaluate the segmentation maps obtained from various sequences using different segmentation approaches.Publication Metadata only The triangle intersection problem for K4 - E designs(Utilitas Mathematica Publishing Inc., 2007) Billington, Elizabeth J.; Lindner, C. C.; Department of Mathematics; Yazıcı, Emine Şule; Faculty Member; Department of Mathematics; College of Sciences; 27432An edge-disjoint decomposition of the complete graph Kn into copies of K4 - e, the simple graph with four vertices and five edges, is known to exist if and only if n ≡ 0 or 1 (mod 5) and n ≥ 6 (Bermond and Schönheim, Discrete Math. 19 (1997)). The intersection problem for K4 - e designs has also been solved (Billington, M. Gionfriddo and Lindner, J. Statist. Planning Inference 58 (1997)); this problem finds the number of common K4 - e blocks which two K4 - e designs on the same set may have. Here we answer the question: how many common triangles may two K4 - e designs on the same set have? Since it is possible for two K4 - e designs on the same set to have no common K4 - e blocks and yet some positive number of common triangles, this problem is largely independent of the earlier K4 - e intersection result.Publication Metadata only A novel test coverage metric for concurrently-accessed software components (A work-in-progress paper)(Springer-Verlag Berlin, 2006) N/A; Department of Computer Engineering; N/A; Department of Computer Engineering; Department of Computer Engineering; Taşıran, Serdar; Elmas, Tayfun; Bölükbaşı, Güven; Keremoğlu, M. Erkan; Faculty Member; PhD Student; Undergraduate Student; Reseacher; Department of Computer Engineering; College of Engineering; Graduate School of Sciences and Engineering; College of Engineering, College of Engineering; N/A; N/A; N/A; N/AWe propose a novel, practical coverage metric called "location pairs" (LP) for concurrently-accessed software components. The LP metric captures well common concurrency errors that lead to atomicity or refinement violations. We describe a software tool for measuring LP coverage and outline an inexpensive application of predicate abstraction and model checking for ruling out infeasible coverage targets.Publication Metadata only Robust speech recognition using adaptively denoised wavelet coefficients(IEEE, 2004) Department of Electrical and Electronics Engineering; Department of Electrical and Electronics Engineering; N/A; Tekalp, Ahmet Murat; Erzin, Engin; Akyol, Emrah; Faculty Member; Faculty Member; Master Student; Department of Electrical and Electronics Engineering; College of Engineering; College of Engineering; Graduate School of Sciences and Engineering; 26207; 34503; N/AThe existence of additive noise affects the performance of speech recognition in real environments. We propose a new set of feature vectors for robust speech recognition using denoised wavelet coefficients. The use of wavelet coefficients in speech processing is motivated by the ability of the wavelet transform to capture both time and frequency information and the non-stationary behaviour of speech signals. We use one set of noisy data, such as data with car noise, and we use hard thresholding in the best basis for denoising. We use isolated digits as our database in our HMM based speech recognition system. A performance comparison of hard thresholding denoised wavelet coefficients and MFCC feature vectors is presented.Publication Metadata only Sculpture surface machining: a generalized model of ball-end milling force system(Elsevier Sci Ltd, 2003) N/A; Department of Mechanical Engineering; Lazoğlu, İsmail; Faculty Member; Department of Mechanical Engineering; College of Engineering; 179391A new mechanistic model is presented for the prediction of a cutting force system in ball-end milling of sculpture surfaces. The model has the ability to calculate the workpiece/cutter intersection domain automatically for a given cutter location (CL) file, cutter and workpiece geometries. Furthermore, an analytical approach is used to determine the instantaneous chip load (with and without runout) and cutting forces. In addition to predicting the cutting forces, the model also employs a Boolean approach for a given cutter, workpiece geometries, and CL file to determine the surface topography and scallop height variations alone, the workpiece surface which can be visualized in 3-D. The results of model validation experiments on machining Ti-6A1-4V are also reported. Comparisons of the predicted and measured forces as well as surface topography show good agreement.Publication Metadata only Value creation in service delivery: relating market segmentation, incentives, and operational performance(Institute for Operations Research and the Management Sciences, 2004) Güneş, Evrim D.; Department of Business Administration; Karaesmen, Zeynep Akşin; Faculty Member; Department of Business Administration; College of Administrative Sciences and Economics; 4534This paper studies service-delivery design in settings where firms engage in value-creation activities that have the objective of generating additional revenue from customer interactions. The paper provides a general modelling framework to analyze the ties between market segmentation decisions, incentives, and process performance in such service-delivery systems. The firm is modelled as a single-server queue, in a principal-agent framework. Customers have different value-generation potentials whose realizations are observed by the server but not by the manager of the firm. The manager determines a market segmentation scheme given an overall customer value-generation profile, which divides customers into two groups (high and low), and also determines a service level for each segment. The server decides which of the two available service levels (high and low) to provide for each customer, given a compensation scheme offered by the manager. The optimal market segmentation decision, optimal service-level choice, and a set of optimal linear incentive contracts that enable their implementation are characterized. The robustness of these strategies is explored with respect to model parameters and assumptions. It is shown that a market segmentation scheme that combines revenue generation concerns with their process implications is essential for success. Characteristics of appropriate incentive schemes are identified.Publication Metadata only Linguistic strategies serving evaluative functions: a comparison between Japanese and Turkish narratives(Lawrence Erlbaum Associates. Inc., 2003) Nakamura, Keiko; Department of Psychology; Küntay, Aylin C.; Faculty Member; Department of Psychology; College of Social Sciences and Humanities; 178879N/APublication Metadata only A computational study of drop formation in an axisymmetric flow-focusing device(Amer Soc Mechanical Engineers, 2006) Department of Mechanical Engineering; Department of Mechanical Engineering; Filiz, İsmail; Muradoğlu, Metin; N/A; Faculty Member; Department of Mechanical Engineering; College of Engineering; College of Engineering; N/A; 46561We investigate the formation and dynamics of drops computationally in an axisymetric geometry using a Front-Tracking/Finite-Difference (FT/FD) method. The effects of viscosity ratio between inner and outer liquids on the drop creation process and drop size distribution are examined. It is found that the viscosity ratio critically influences the drop formation process and the final drop distribution. We found that, for small viscosity ratios, i.e., 0.1 < lambda < 0.5 drop size is about the size of the orifice and drop distribution is highly monodisperse. When viscosity ratio is increased, i.e., 0.5 < lambda < I a smaller drop is created just after the main drop. For even higher viscosity ratios, the drop distribution is usually monodisperse but a satellite drop is created in some cases. The effect of the flow rates in the inner jet and the co flowing annulus are also studied. It is found that the drop size gets smaller as Q(in) / Q(out) is reduced while keeping the outer flow rate constant.