The maintenance of a system on a ship has limitations when the ship is engaged in a voyage because of limited maintenance resources. When a system fails, it is either repaired instantly on ship with probability p or remains unrepaired during the voyage with probability 1-p owing to the lack of maintenance resources. In the latter case, the system is repaired after the voyage. We propose two management policies for the overhaul interval of an IFR system: one manages the overhaul interval by number of voyages and the other manages it by the total voyage time. Our goal is to determine the optimal policy that ensures the required availability of the system and minimizes the expected cost rate.

This paper proposes a Bayesian approach to image recognition based on separable lattice hidden Markov models (SL-HMMs). The geometric variations of the object to be recognized, e.g., size, location, and rotation, are an essential problem in image recognition. SL-HMMs, which have been proposed to reduce the effect of geometric variations, can perform elastic matching both horizontally and vertically. This makes it possible to model not only invariances to the size and location of the object but also nonlinear warping in both dimensions. The maximum likelihood (ML) method has been used in training SL-HMMs. However, in some image recognition tasks, it is difficult to acquire sufficient training data, and the ML method suffers from the over-fitting problem when there is insufficient training data. This study aims to accurately estimate SL-HMMs using the maximum a posteriori (MAP) and variational Bayesian (VB) methods. The MAP and VB methods can utilize prior distributions representing useful prior information, and the VB method is expected to obtain high generalization ability by marginalization of model parameters. Furthermore, to overcome the local maximum problem in the MAP and VB methods, the deterministic annealing expectation maximization algorithm is applied for training SL-HMMs. Face recognition experiments performed on the XM2VTS database indicated that the proposed method offers significantly improved image recognition performance. Additionally, comparative experiment results showed that the proposed method was more robust to geometric variations than convolutional neural networks.

Embedded memory is extensively being used in SoCs, and is rapidly growing in size and density. It contributes to SoCs to have greater features, but at the expense of taking up the most area. Due to continuous scaling of nanoscale device technology, large area size memory introduces aging-induced faults and soft errors, which affects reliability. In-field test and repair, as well as ECC, can be used to maintain reliability, and recently, these methods are used together to form a combined approach, wherein uncorrectable words are repaired, while correctable words are left to the ECC. In this paper, we propose a novel in-field repair strategy that repairs uncorrectable words, and possibly correctable words, for an ECC-based memory architecture. It executes an adaptive reconfiguration method that ensures 'fresh' memory words are always used until spare words run out. Experimental results demonstrate that our strategy enhances reliability, and the area overhead contribution is small.

As three dimensional (3D) discrete wavelet transform (DWT) is widely used for high resolution volumetric data compression, and to further improve the performance of lossless coding, the adaptive directional lifting (ADL) structure based on non-separable 3D DWT with a (5,3) filter is proposed in this paper. The proposed 3D DWT has less lifting steps and better prediction performance compared to the existing separable 3D DWT with fixed filter coefficients. It also has compatibility with the conventional DWT defined by the JPEG2000 international standard. The proposed method shows comparable and better results with the non-separable 3D DWT and separable 3D DWT and it is effective for lossless coding of high resolution volumetric data.

The information maximization (Infomax) based on information entropy theory is a class of methods that can be used to blindly separate the sources. Torkkola applied the Infomax criterion to blindly separate the mixtures where the sources have been delayed with respect to each other. Compared to the frequency domain methods, this time domain method has simple adaptation rules and can be easily implemented. However, Torkkola's method works only in the real valued field. In this letter, the Infomax for blind separation of the delayed sources is extended to the complex case for processing of complex valued signals. Firstly, based on the gradient ascent the adaptation rules for the parameters of the unmixing network are derived and the steps of algorithm are given. Then, a measurement matrix is constructed to evaluate the separation performance. The results of computer experiment support the extended algorithm.

Proof of Retrievability (POR) is a protocol by which a client can distribute his/her data to cloud servers and can check if the data stored in the servers is available and intact. After that, network coding-based POR has been applied to improve network throughput. Although many network coding-based PORs have been proposed, most of them have not achieved the following practical features: direct repair and dynamic operations. In this paper, we propose the D2-POR scheme (Direct repair and Dynamic operations in network coding-based POR) to address these shortcomings. When a server is corrupted, the D2-POR can support the direct repair in which the data stored in the corrupted server can be repaired using the data directly provided by healthy servers. The client is thus free from the burden of data repair. Furthermore, the D2-POR allows the client to efficiently perform dynamic operations, i.e., modification, insertion and deletion.

In distributed storage systems, codes with lower repair locality are much more desirable due to their superiority in reducing the disk I/O complexity of each repair process. Motivated partially by both codes with information (r,δ1)c locality and codes with cooperative (r,l) locality, we propose the concept of codes with information (r,l,δ) locality in this paper. For a linear code C with information (r,l,δ) locality, values at arbitrary l information coordinates of an information set I can be recovered by connecting any of δ existing pairwise disjoint local repair sets with size no more than r, where a local repair set of l coordinates is defined as the set of some other coordinates by which one can recover the values at these l coordinates. We derive a lower bound on the codeword length n for [n,k,d] linear codes with information (r,l,δ) locality. Furthermore, we indicate its tightness for some special cases. Particularly, some existing results can be deduced from our bound by restriction on parameters.

Cache memories are the major application of high-speed SRAMs, and they are frequently installed in high performance logic VLSIs including microprocessors. This paper presents a 4-way set-associative, SOI cache-tag memory. To obtain higher operating speed with less power dissipation, we devised an I/O-separated memory cell with a dual-rail wordline, which is used to transmit complementary selection signals. The address decoding delay was shortened using CMOS dual-rail logic. To enhance the maximum operating frequency, bitline's recovery operations after writing data were eliminated using a memory array configuration without half-selected cells. Moreover, conventional, sensitive but slow differential amplifiers were successfully removed from the data I/O circuitry with a hierarchical bitline scheme. As regards the stored data management, we devised a new hardware-oriented LRU-data replacement algorithm on the basis of 6-bit directed graph. With the experimental results obtained with a test chip fabricated with a 0.25-µm CMOS/SIMOX process, the core of the cache-tag memory with a 1024-set configuration can achieve a 1.5-ns address access time under typical conditions of a 2-V power supply and 25°C. The power dissipation during standby was less than 14 µW, and that at the 500-MHz operation was 13-83 mW, depending on the bit-stream data pattern.

Intelligent information systems captivate people's attention. Examples of such systems include driving support vehicles capable of sensing driver state and communication robots capable of interacting with humans. Modeling how people search visual information is indispensable for designing these kinds of systems. In this paper, we focus on human visual attention, which is closely related to visual search behavior. We propose a computational model to estimate human visual attention while carrying out a visual target search task. Existing models estimate visual attention using the ratio between a representative value of visual feature of a target stimulus and that of distractors or background. The models, however, can not often achieve a better performance for difficult search tasks that require a sequentially spotlighting process. For such tasks, the linear separability effect of a visual feature distribution should be considered. Hence, we introduce this effect to spatially localized activation. Concretely, our top-down model estimates target-specific visual attention using Fisher's variance ratio between a visual feature distribution of a local region in the field of view and that of a target stimulus. We confirm the effectiveness of our computational model through a visual search experiment.

This paper proposes fast repairing methods that uses hierarchical software defined network controllers for recovering from massive failure in a large-scale IP over a wavelength-division multiplexing network. The network consists of multiple domains, and slave controllers are deployed in each domain. While each slave controller configures transport paths in its domain, the master controller manages end-to-end paths, which are established across multiple domains. For fast repair of intra-domain paths by the slave controllers, we define the optimization problem of path configuration order and propose a heuristic method, which minimizes the repair time to move from a disrupted state to a suboptimal state. For fast repair of end-to-end path through multiple domains, we also propose a network abstraction method, which efficiently manages the entire network. Evaluation results suggest that fast repair within a few minutes can be achieved by applying the proposed methods to the repairing scenario, where multiple links and nodes fail, in a 10,000-node network.

This letter presents a two-stage method to extend the degenerate unmixing estimation technique (DUET) for reverberant speech separation. First, frequency-bin-wise attenuation and delay parameters are introduced and estimated by online update rules, to handle early reflections. Next, a mask reestimation algorithm based on the precedence effect is developed to detect and fix the errors on binary masks caused by late reflections. Experimental results demonstrate that the proposed method improves separation performance significantly.

A one-shot system is a system that can be used only once during its life, and whose failures are detected only through inspections. In this paper, we discuss an inspection policy problem of one-shot system composed of multi-unit in series. Failed units are minimally repaired when failures are detected and all units in the system are replaced when the nth failure is detected after the last replacement. We derive the expected cost rate approximately. Our goal is to determine the optimal inspection policy that minimizes the expected cost rate.

Fault localization is essential for conducting effective program repair. However, preliminary studies have shown that existing fault localization approaches do not take the requirements of automatic repair into account, and therefore restrict the repair performance. To address this issue, this paper presents the first study on designing fault localization approaches for automatic program repair, that is, we propose a fault localization approach using failure-related contexts in order to improve automatic program repair. The proposed approach first utilizes program slicing technique to construct a failure-related context, then evaluates the suspiciousness of each element in this context, and finally transfers the result of evaluation to automatic program repair techniques for performing repair on faulty programs. The experimental results demonstrate that the proposed approach is effective to improve automatic repair performance.

In this paper, we present update rules for convolutive nonnegative matrix factorization (NMF) in which cost functions are based on the squared Euclidean distance, the Kullback-Leibler (KL) divergence and the Itakura-Saito (IS) divergence. We define an auxiliary function for each cost function and derive the update rule. We also apply this method to the single-channel signal separation in speech signals. Experimental results showed that the convergence of our KL divergence-based method was better than that in the conventional method, and our method achieved single-channel signal separation successfully.

The goal of dimension reduction is to represent high-dimensional data in a lower-dimensional subspace, while intrinsic properties of the original data are kept as much as possible. An important challenge in unsupervised dimension reduction is the choice of tuning parameters, because no supervised information is available and thus parameter selection tends to be subjective and heuristic. In this paper, we propose an information-theoretic approach to unsupervised dimension reduction that allows objective tuning parameter selection. We employ quadratic mutual information (QMI) as our information measure, which is known to be less sensitive to outliers than ordinary mutual information, and QMI is estimated analytically by a least-squares method in a computationally efficient way. Then, we provide an eigenvector-based efficient implementation for performing unsupervised dimension reduction based on the QMI estimator. The usefulness of the proposed method is demonstrated through experiments.

Since electromagnetic (EM) noise resulting from an arc discharge disturbs other electric devices, parameters on electromagnetic compatibility, as well as lifetime and reliability, are important properties for electrical contacts. To clarify the characteristics and the mechanism of the generation of the EM noise, the arc column and voltage fluctuations generated by slowly breaking contacts with external direct current (DC) magnetic field, up to 20,mT, was investigated experimentally using Ag$_{90.7{ m wt%}}$SnO$_{2,9.3{ m wt}%}$ material. Firstly the motion of the arc column is measured by high-speed camera. Secondary, the distribution of the motion of the arc and contact voltage are discussed. It was revealed that the contact voltage fluctuation in the arc duration is related to the arc column motion.

To study the molten bridge phenomenon of contacts at the initial breaking process, an experimental device of molten bridge between slowly opening contacts was developed. The system consists of the contact moving control module, the circuit load and the observation module. The molten bridge of copper contact under two load conditions 9,V/19,A and 9,V/7.3,A were studied. The voltage and current characteristics curves of Cu molten bridge were extracted and the resistance and the instantaneous power of the molten bridge were analyzed. The image of the Cu molten bridge diameter was captured by CCD under 9,V/19,A and the influences of the contact force and the separation speed on the molten bridge length and the crater diameter of the anode were studied. The root profile of the Cu contacts after separation was analyzed by digital microscope. Research results show that the Cu molten bridge length has the same changing trend as the diameter of the anode crater. They both decrease with the increment of the separation speed and the decrement of the contact force.

This paper proposes a so called quasi-linear support vector machine (SVM), which is an SVM with a composite quasi-linear kernel. In the quasi-linear SVM model, the nonlinear separation hyperplane is approximated by multiple local linear models with interpolation. Instead of building multiple local SVM models separately, the quasi-linear SVM realizes the multi local linear model approach in the kernel level. That is, it is built exactly in the same way as a single SVM model, by composing a quasi-linear kernel. A guided partitioning method is proposed to obtain the local partitions for the composition of quasi-linear kernel function. Experiment results on artificial data and benchmark datasets show that the proposed method is effective and improves classification performances.

This paper reduces the total number of lifting steps of a two dimensional (2D) discrete wavelet transform (DWT) under the constraint that the DWT has full compatibility with a given separable 2D DWT. In a DWT composed of several lifting steps in cascade, a lifting step must wait for a calculation result of its previous lifting step. Therefore more lifting steps bring about longer latency from the input to the output. In this paper, we reduce the total number of lifting steps of a quadruple lifting DWT which is widely utilized as the 9/7 DWT in the JPEG 2000 international standard for image data compression. Firstly, we introduce basic properties for modification of signal flow structure inside a DWT without changing its output. Secondly, we convert the separable 2D quadruple lifting DWT into the non-separable 2D DWT utilizing the basic properties so that the total number of the lifting steps is reduced. As a result, the lifting steps were reduced to 75[%]. Finally, we evaluate the proposed non-separable 2D DWT in respect of compatibility with the conventional separable 2D DWT.

In this paper, a novel statistical model based on 2-D HMMs for image recognition is proposed. Recently, separable lattice 2-D HMMs (SL2D-HMMs) were proposed to model invariance to size and location deformation. However, their modeling accuracy is still insufficient because of the following two assumptions, which are inherited from 1-D HMMs: i) the stationary statistics within each state and ii) the conditional independent assumption of state output probabilities. To overcome these shortcomings in 1-D HMMs, trajectory HMMs were proposed and successfully applied to speech recognition and speech synthesis. This paper derives 2-D trajectory HMMs by reformulating the likelihood of SL2D-HMMs through the imposition of explicit relationships between static and dynamic features. The proposed model can efficiently capture dependencies between adjacent observations without increasing the number of model parameters. The effectiveness of the proposed model was evaluated in face recognition experiments on the XM2VTS database.