Support of collaborative work and management of spatio-temporal data has become one of the most interesting and important database applications, which is due to the tremendous progress of database and its surrounding technologies in the last decade. In this paper, we investigate the new generation database technologies that are needed to support such advanced applications. Because of the recent progress of virtual reality technology, virtual work spaces are now available. We examine a typical CSCW (Computer Supported Cooperative Work) fsystem to identify database problems that arise from it. We introduce typical approaches to database improvement based on the high-level view and the virtual reality technique. Also, in this paper, the following are introduced and discussed: the design and implementation of three- and four-dimensional spatio-temporal database systems, VRML (Virtual Reality Modeling Language) database systems, fast access methods to spatio-temporal data, and the interval-based approach to temporal multimedia databases.

This paper presents a new efficient method for finding an "optimal" bi-decomposition form of a logic function. A bi-decomposition form of a logic function is the form: f(X) = α(g1(X1), g2(X2)). We call a bi-decomposition form optimal when the total number of variables in X1 and X2 is the smallest among all bi-decomposition forms of f. This meaning of optimal is adequate especially for the synthesis of LUT (Look-Up Table) networks where the number of function inputs is important for the implementation. In our method, we consider only two bi-decomposition forms; (g1 g2) and (g1 g2). We can easily find all the other types of bi-decomposition forms from the above two decomposition forms. Our method efficiently finds one of the existing optimal bi-decomposition forms based on a branch-and-bound algorithm. Moreover, our method can also decompose incompletely specified functions. Experimental results show that we can construct better networks by using optimal bi-decompositions than by using conventional decompositions.

Simple disjunctive decomposition is a special case of logic function decompositions, where variables are divided into two disjoint sets and there is only one newly introduced variable. It offers an optimal structure for a single-output function. This paper presents two techniques that enable us to apply simple disjunctive decompositions with little overhead. Firstly, we propose a method to find symple disjunctive decomposition forms efficiently by limiting decomposition types to be found to two: a decomposition where the bound set is a set of symmetric variables and a decomposition where the output function is a 2-input function. Secondly, we propose an algorithm that constructs a new logic representation for a simple disjunctive decomposition just by assigning constant values to variables in the original representation. The algorithm enables us to apply the decomposition with keeping good structures of the original representation. We performed experiments for decomposing functions and confirmed the efficiency of our method. We also performed experiments for restructuring fanout free cones of multi-level logic circuits, and obtained better results than when not restructuring them.

In this paper, we analyze the inverse scattering problem by a new deterministic method called "Source and Radiation Field Solution," which has the merit that both the source and the radiation field can be treated at the same time, the effect of which has already shown in ordinary scattering problems.

Providing various assistances for design modifications on HDL source codes is important for design reuse and quick design cycle in VLSI CAD. Program slicing is a software-engineering technique for analyzing, abstracting, and transforming programs. We show algorithms for extracting/removing behaviors of specified signals in VHDL descriptions. We also describe a VHDL slicing system and show experimental results of efficiently extracting components from VHDL descriptions.

In the task of forming high-level object-centered models from low-level image-based features, parts serve as an intermediate representation. A representation of parts for object recognition should be rich, stable, and invariant to changes in the viewing conditions. In addition, it should be capable of describing partially occluded shapes. This paper describes a method for decomposing shapes into parts. The method is based on pairs of negative curvature minima which have a good continuation at their boundary tangents. A measure of good continuation is proposed by using the coefficients of cocircularity, smoothness, and proximity. This method could recover parts in a direct computation, therefore efficient in calculation than the former. Currently, we assume that the shape is a closed planar curve.

This paper describes a new Artificial Neural Network (ANN), UNItary Decomposition ANN (UNIDANN), which can perform the unitary eigendecomposition of the synaptic weight matrix. It is shown both analytically and quantitatively that if the synaptic weight matrix is Hermitian positive definite, the neural output, based on the proposed dynamic equation, will converge to the principal eigenvectors of the synaptic weight matrix. Compared with previous works, the UNIDANN possesses several advantageous features such as low computation time and no synchronization problem due to the underlying analog circuit structure, faster convergence speed, accurate final results, and numerical stability. Some simulations with a particular emphasis on the applications to high resolution bearing estimation problems are also furnished to justify the proposed ANN.

The DC component suppressing method, called Guided Scrambling (GS), has been proposed, where a source bit stream within a data block is subjected to several kinds of scrambling and a RLL (Run Length Limited) coding to make the selection set of channel bit streams, then the one having the least DC component is selected. Typically, this technique uses a convolutional operation or GF (Galois field) conversion. A review of their respective symbol error properties has revealed important findings. In the former case, the RS (Reed-Solomon) decoding capability is reduced because error propagation occurs in descrambling. In the latter case, error propagation of a data block length occurs when erroneous conversion data occurs after RS decoding. This paper introduces expressions for determining the decoded symbol error probabilities of the two schemes based on these properties. The paper also discusses the difference in code rates between the two schemes on the basis of the result of calculation using such expressions.

This study shows the effectiveness of the simulation probability density function (p. d. f. ) based on the Bhattacharyya bound from the point of view of the twisted distribution. As a result, the simulation p. d. f. related to the Bhattacharyya bound is asymptotically optimal for the trellis coded modulation scheme under some practical conditions. And the optimality is also confirmed by a numerical example.

An asynchronous pulse neural network model which is suitable for VLSI implementation is proposed. The model neuron can function as a coincidence detector as well as an integrator depending on its internal time-constant relative to the external one, and show complex dynamical behavior including chaotic responses. A network with the proposed neurons can process spatio-temporal coded information through dynamical cell assemblies with functional synaptic connections.

This paper describes an efficient simulation algorithm for the spatiotemporal pattern analysis of the continuous-time neural networks with the multivalued logic (multivalued continuous-time neural networks). The multivalued transfer function of neuron is approximated to the stepwise constant function which is constructed by the sum of the step functions with the different thresholds. By this approximation, the dynamics of the network can be formulated as a stepwise constant linear differential equation at each timestep and the optimal timestep for the numerical integration can be obtained analytically. Finally, it is shown that the proposed method is much faster than a variety of conventional simulators.

In this study, a coupled chaotic circuits network is realized by real circuit elements. By using a simple circuit converting generating spatial patterns to digital signal, irregular self-switching phenomenon of the appearing patterns can be observed as real physical phenomenon.

Simplification orderings, like the recursive path ordering and the improved recursive decomposition ordering, are widely used for proving the termination property of term rewriting systems. The improved recursive decomposition ordering is known as the most powerful simplification ordering. Recently Jouannaud and Rubio extended the recursive path ordering to higher-order rewrite systems by introducing an ordering on type structure. In this paper we extend the improved recursive decomposition ordering for proving termination of higher-order rewrite systems. The key idea of our ordering is a new concept of pseudo-terminal occurrences.

In this paper, the performances of texture classification based on pyramidal and uniform decomposition are comparatively studied with and without feature selection. This comparison using the subband variance as feature explores the dependence among features. It is shown that the main problem when employing 2-D non-separable wavelet transforms for texture classification is the determination of the suitable features that yields the best classification results. A Max-Max algorithm which is a novel evaluation function based on genetic algorithms is presented to evaluate the classification performance of each subset of selected features. It is shown that the performance with feature selection in which only about half of features are selected is comparable to that without feature selection. Moreover, the discriminatory characteristics of texture spread more in low-pass bands and the features extracted from the pyramidal decomposition are more representative than those from the uniform decomposition. Experimental results have verified the selectivity of the proposed approach and its texture capturing characteristics.

A novel approach was proposed to recognize the non-rigid 3D objects from their corresponding 2D images by combining the benefits of the principal component analysis and the geometric hashing. For all of the object models to be recognized, we calculated the statistical point features of the training shapes using principal component analysis. The results of the analysis were a vector of eigenvalues and a matrix of eigenvectors. We calculated invariants of the new shapes that undergone a similarity transformation. Then added these invariants and the label of the model to the model database. To recognize objects, we calculated the necessary invariants from an unknown image and used them as the indexing keys to retrieve any possible matches with the model features from the model database. We hypothesized the existence of an instance of the model in the scene if the model's features scored enough hits on the vote count. This approach allowed us to store the rigid and the non-rigid object models in a model database and utilized them to recognize an instance of model from an unknown image.

In a variety of communication systems, the multipath propagation due to various reflections is often encountered. In this paper, the directions-of-arrival (DOA) estimation of the cyclostationary coherent signals is investigated. A new approach is proposed for estimating the DOA of the coherent signals impinging on a uniform linear array (ULA) by utilizing the spatial smoothing (SS) technique. In order to improve the robustness of the DOA estimation by exploiting the cyclic statistical information sufficiently and handling the coherence effectively, we give a cyclic algorithm with multiple lag parameters and the optimal subarray size. The performance of the presented method is verified and compared with the conventional methods through numerical examples.

In multilevel block modulation codes for QPSK and 8-PSK modulation, a construction of binary component codes is given. These codes have a good minimum Euclidean distance by using different forms of the dependency properties of the binary component codes. Interdependency among component codes is formed by using the binary component subcodes which are derived by the coset decomposition of the binary component codes. The algebraic structures of the codes are investigated to find out how interdependency among component codes gives a good minimum Euclidean distance. First, it is shown that cyclic codes over ZM for M-PSK (M=4,8), where the coding scheme is given by Piret, can be constructed by forming specific interdependency among binary component codes for proposed multilevel coding method. Furthermore, it is shown that better minimum Euclidean distance than above can be obtained by modifying the composition of interdependency among binary component codes. These proposed multilevel codes have algebraic structure of additive group and cyclic property over GF(M). Finally, error performances are compared with those of some code's reference modulation scheme for transmitting the same number of information bits.

We introduce a strongly infix code. A code X is a strongly infix code if X is an infix code and any catenation of two words in X has no proper factor in X, which is neither a left factor nor a right factor. We show that the class of strongly infix codes is closed under composition, and, as the dual result, that the property to be strongly infix is inherited by a component of a decomposition.

Huynen has already provided a method to decompose a Mueller matrix in order to retrieve detailed target information in a polarimetric radar system. However, this decomposition sometimes fails in the presence of small error or noise in the elements of a Mueller matrix. This paper attempts to improve Huynen's decomposition method. First, we give the definition of stable decomposition and present an example, showing a problem of Huynen's approach. Then two methods are proposed to carry out stable decompositions, based on the nonlinear least square method and the Newton's method. Stability means the decomposition is not sensitive to noise. The proposed methods overcomes the problems on the unstable decomposition of Mueller matrix, and provides correct information of a target.

We present experimental results for a structural learning method of feed-forward neural-network classifiers using Principal Component Analysis (PCA) network and Species Genetic Algorithm (SGA). PCA network is used as a means for reducing the number of input units. SGA, a modified GA, is employed for selecting the proper number of hidden units and optimizing the connection links. Experimental results show that the proposed method is a useful tool for choosing an appropriate architecture for high dimensions.