This paper presents a method for spatial and temporal segmentation of long image sequences which include multiple independently moving objects, based on the Minimum Description Length (MDL) principle. By obtaining an optimal motion description, we extract spatiotemporal (ST) segments in the image sequence, each of which consists of edge segments with similar motions. First, we construct a family of 2D motion models, each of which is completely determined by its specified set of equations. Then, based on these sets of equations we formulate the motion description length in a long sequence. The motion state of one object at one moment is determined by finding the model with shortest description length. Temporal segmentation is carried out when the motion state is found to have changed. At the same time, the spatial segmentation is globally optimized in such a way that the motion description of the entire scene reaches a minimum.

Over the last decade, research in the automatic synthesis and optimization of combinational logic has matured significantly; more recently, research has focused on sequential logic. Many of the paradigms for combinational logic have been extended and applied in the sequential domain. In addition, promising new directions for future research are being explored. In this paper, we survey some of the results of combinational synthesis and some recent results for sequential synthesis and then use these to view possible avenues for future sequential synthesis research. In particular we look at two related questions: deriving a set of permissible behaviors and using a minimizer to select the best behavior according to some optimization criteria. We examine these two issues in increasingly complex situations starting with a single-output function, and proceeding to a single multiple-output function, a network of single-output functions, a network of multiple-output functions, and then similar questions where function" is replaced by a finite state machine (FSM). We end with a discussion of a network of finite state machines and the problem of deriving the set of permissible FSM's and choosing a representative minimum one.

This paper investigates the problem of scheduling parallel tasks" with consideration of communication cost on an m-processor system, where processors are assumed to be identical and tasks being scheduled are independent such that they can run on more than one processor simultaneously. Once a task is processed in parallel, its finish time will be speeded up, but communication cost will also be incurred and should be taken into account. To find a schedule with minimum finish time for the parallel tasks scheduling problem is NP-hard. Therefore, in this paper, we will propose a heuristic algorithm for this kind of problem and derive its performance bounds for two different cases of applications, respectively.

In this paper we introduce a new feature called stroke index for document image analysis. It is based on the minimum covering run expression method (MCR). This stroke index is a function of the number of horizontal and vertical runs in the original image and of number of runs by the MCR expression. As document images may present a variety of patterns such as graph, text or picture, it is necessary for image understanding to classify these different patterns into categories beforehand. Here we show how one could use this stroke index for such applications as classification or segmentation. It also gives an insight on the possibility of stroke extraction from document images in addition to classifying different patterns in a compound image.

A scalable and flexible ASIC CIM system distributed on UNIX workstations, ORCHARD , has been developed. It is designed from three viewpoints: (1) cost and TAT reduction in system construction, (2) flexibility in data management for quality control, and (3) precise and quick scheduling and effective lot tracking to control TAT for each lot. The concept of a "virtual machine" is introduced to connect equipment having various protocols to a host system. The virtual machine is automatically generated at an average automatic generation ratio of as high as 89%, which leads to a reduction in cost and TAT in system construction. Data for quality control is managed by changing flexibly the "data processing recipe." This recipe defines screen format, data collected from equipment, and data transfered from various databases. Precise scheduling of lots with various levels of priority is achieved by introducing a priority evaluation function, thereby reducing scheduling time to 1/20 that for manual scheduling.

Micro-tribology is a key technology in micro-machine. Atomic-scale wear and friction fluctuations degrade the performance of micro-machines. New wear-resistant, low friction materials should be useful in reducing micro- and macro-tribological wear and friction fluctuations. Our investigation of the frictional characteristics of polished CVD diamond films by FFM (friction force microscope), AFM (atomic force microscope) and conventional reciprocating tribometer and trial micro processing of diamond produced three main results. First, the friction coefficient of diamond film increases rapidly with decreasing load in the micro-load region. This is partially due to the surface tension of adsorbed water on the surface under high humidity. In the macro-load region also, the friction coefficient increases with decreasing load, but, in this case it is due to elastic deformation. The second result is that diamond film has excellent wear resistance in the micro-load region compared with silicon and diamond-like carbon (DLC) film. Finally, a micro-diamond gear and diamond shaft were fabricated by laser machining and thermo-chemical etching, and then assembled.

In the present paper, we attempt to show that the information about input patterns must be as small as possible for improving the generalization performance under the condition that the network can produce targets with appropriate accuracy. The information is defined with respect to the hidden unit activity and we suppose that the hidden unit has a crucial role to store the information content about input patterns. The information is defined by the difference between uncertainty of the hidden unit at the initial stage of the learning and the uncertainty of the hidden unit at the final stage of the learning. After having formulated an update rule for the information minimization, we applied the method to a problem of language acquisition: the inference of the past tense forms of regular and irregular verbs. Experimental results confirmed that by our method, the information was significantly decreased and the generalization performance was greatly improved.

Dynamic Term Rewriting Calculus (DTRC) is a new computation model aiming at formal description and verification of algorithms treating Term Rewriting Systems (TRSs). In this paper, we show that we can use DTRC to mechanize explicit induction for proving an inductive theorem, that is, we can translate the statements of base and induction steps for proving by induction into a DTRC term. The translation reduces the proof of the statements into the evaluation of the corresponding DTRC term.

This paper describes an advanced rule-embedded neural network (RENN+) that has an extended framework for achieving a very tight integration of learning-based neural networks and rule-bases of existing if-then rules. The RENN+ is effective in pattern recognition with ill-posed conditions. It is basically composed of several component RENNs and an output RENN, which are three-layer back-propagation (BP) networks except for the input layer. Each RENN can be pre-organized by embedding the if-then rules through translation of the rules into logic functions in a disjunctive normal form, and can be trainded to acquire adaptive rules as required. A weight-modification-reduced learning algorithm (WMR) capable of standard regularization is used for the post-training to suppress excessive modification of the weights for the embedded rules. To estimate the effectiveness of the proposed RENN+, it was used for pattern recognition in a radar system for detection of buried pipes. This trial showed that a RENN+ with two component RENNs had good recognition capability, whereas a conventional BP network was ineffective.

We define a new framework for rewriting graphs, called a formal graph system (FGS), which is a logic program having hypergraphs instead of terms in first-order logic. We first prove that a class of graphs is generated by a hyperedge replacement grammar if and only if it is defined by an FGS of a special form called a regular FGS. In the same way as logic programs, we can define a refutation tree for an FGS. The classes of TTSP graphs and outerplanar graphs are definable by regular FGSs. Then, we consider the problem of constructing a refutation tree of a graph for these FGSs. For the FGS defining TTSP graphs, we present a refutation tree algorithm of O(log2nlogm) time with O(nm) processors on an EREW PRAM. For the FGS defining outerplanar graphs, we show that the refutation tree problem can be solved in O(log2n) time with O(nm) processors on an EREW PRAM. Here, n and m are the numbers of vertices and edges of an input graph, respectively.

The efficiency of a guided-probe fault location process is affected by the number of the probed lines. This number depends on the size of the target area and the method by which a line is selected for probing. This paper presents a method for reducing the size of the target area in a sequential circuit by introducing the concepts of Type- and Type- faults. This paper also presents a method of selecting lines for probing in a more efficient way. The efficiency of the proposed methods is demonstrated by experimental results.

In this paper, a design of TCM signals for Middleton's class-A impulsive noise environment is investigated. The error event characteristics under the impulsive noise is investigated, and it is shown that the length of the signal sequence is more important than Euclidean distance between the signal sequences. Following this fact, we introduce the shortest error event path length as a measure of the signal design. In order to make this value large, increasing of states of convolutional codes is employed, and the performance improvement achieved by this method is evaluated. Numerical results show the great improvement of the error performance and conclude that the shortest error event path length is a good measure in the design of TCM signals under impulsive noise environment. Moreover, the capacity of class-A impulsive noise channel is evaluated, and the required signal sets expansion rates to obtain the achievable coding gain is discussed.

In this letter, we demonstrate an experimental CMOS neural circuit towards an understanding of how particular computations can be performed by a T-Model neural network. The architecture and a digital hardware implementation of the learning T-Model network are presented. Our experimental results show that the T-Model allows immense collective network computations and powerful learning.

In this paper, a new off-line handwritten word recognition method based on the explicit modeling of character junctures is presented. A handwritten word is regarded as a sequence of characters and junctures of four types. Hence both characters and junctures are explicitly modeled. A handwriting system employing hidden Markov models as the main statistical framework has been developed based on this scheme. An interconnection network of character and ligature models is constructed to model words of indefinite length. This model can ideally describe any form of hamdwritten words including discretely spaced words, pure cursive words, and unconstrained words of mixed styles. Also presented are efficient encoding and decoding schemes suitable for this model. The system has shown encouraging performance with a standard USPS database.

A permutation cipher scheme using polynomials over a field is presented. A permutation as well as substitution plays a major role in almost all conventional cryptosystems. But the security of the permutation depends on how symbols are permuted. This paper proposes the use of polynomials for the permutation and show that the scheme satisfies the following security criteria. (1) There are enough encryption keys to defend exhaustive attacks. (2) The permutation moves almost all samples into places which are different from the original places. (3) Most samples are shifted differently by different permutations. The permutation cipher scheme could be regarded as a scheme based on Reed-Solomon codes. The information symbols of the codes compose a key of the permutation cipher scheme.

This paper reports on a new dynamic scanning force microscope (SFM), in which the piezoelectric microcantilever is utilized for the lever excitation and displacement sensing. Piezoelectric cantilevers can detect their deflection without external sensing elements and be vibrated with no oscillator outside. The cantilever integrated with the deflection detector and the oscillator changes the conventional construction of a dynamic SFM and expands its range of applicability. The microcantilever used consists of a ZnO layer sandwiched with Au electrodes deposited on a thin beam of thermally grown SiO2. The length, width and thickness of the lever are 125 µm, 50 µm and 3.5 µm, respectively. We have characterized this cantilever by measuring the charge spectrum and the frequency dependence of the admittance. From the charge spectrum the mechanical quality factor measured 300 in free vibration. Typical piezoelectric constant of the ZnO film was estimated approximately as 80% of single-crystal's value. The piezoelectric cantilever can be vibrated by applying the voltage with the frequency near the resonance to the piezoelectric layer. The excited amplitude per unit voltage at the resonance frequency was calculated as about 5 µm/V. The cantilever amplitude can be detected by measuring the current between electrodes, since the admittance depends on the quality factor. We have constructed a dynamic SFM without external oscillator and detector, and successfully obtained the surface images of a sol-gel derived PZT film in the cyclic contact operation mode. The longitudinal resolution of the SFM system was 0.3 nm at a 125 Hz bandwidth.

Effects of various joint configurations and gap sizes on the electromagnetic shielding effectiveness (SE) are evaluated to provide guidelines for best design of joints in order to increase the value of SE. Four different joint geometries are presented. A sharp decrease on SE with larger gap size for simple joints is observed. Addition of bends in the joint geometry has strong positive effect on the value of the SE. Increasing the angle of cut, which increases the effective length of the joint were also demonstrated to have increasing effect on the shielding performance.

It is shown by the derivation of solution methods for an elementary optimization problem that the stochastic relaxation in image analysis, the Potts neural networks for combinatorial optimization and interior point methods for nonlinear programming have common formulation of their dynamics. This unification of these algorithms leads us to possibility for real time solution of these problems with common analog electronic circuits.

An excitation signal for a synthesis filter plays an important role in producing high quality speech at a low bit rate. This paper presents a new efficient excitation model, Adaptive Density Pulse (ADP) , for low bit-rate speech coding. This ADP is a pulse train whose density (spacing interval) is constant within a subframe but can be varied subframe by subframe. First, the ADP excitation signal is defined. A procedure for finding the optimal ADP excitation is presented. Some results on investigating the effects of the ADP parameters on the synthesized speech quality are discussed. ADP excitation is introduced to the CELP (Code Excited Linear Prediction) coding method to improve speech quality at bit rates around 4 kbps. A CELP coder with an ADP (ADP-CELP) is described. ADP excitation makes it possible for the CELP coder to follow transient portions of speech signals. Also ADP excitation can reduce computational complexity in selecting the best excitation from a codebook, which has been the primary drawback of CELP. The number of multiplications can be reduced to the order of 1/D2 by utilizing the sparseness of ADP excitation, where D is the pulse interval. The authors evaluated the speech quality of a 4 kbps ADP-CELP coder by computer simulation. ADP excitation improved the performance of conventional CELP in segmental SNR.

This paper proposes the overlapped block relaxation Newton method for greatly reducing the number of iterations needed for simulating large scale nonlinear circuits. The circuit is partitioned into subcircuits, i.e., overlapped blocks consisting of core nodes and overlapped nodes. The core nodes form the core circuit for each overlapped block and the overlapped nodes form the overlapped circuit. The Newton-Raphson method is applied to all overlapped blocks independently and the approximation vector for relaxation is determined by node voltages of core nodes. An overlapped circuit is considered to be the representative circuit of the outside circuit for the core circuit. Therefore, the accuracy of the approximation vector for relaxation may be improved and the number of relaxation steps may be greatly reduced. Core nodes are determined automatically by reflecting the circuit structure, then the overlapping level is determined automatically. We show that this method has good performance for simulating large scale circuits, and that it is faster than the nonlinear direct method which is used in standard circuit simulators.