After exploring the classic Lambertian reflectance model, we proposed an effective illumination estimation model to extract illumination invariants for face recognition under complex illumination conditions in this paper. The estimated illumination by our method not only meets the actual lighting conditions of facial images, but also conforms to the imaging principle. Experimental results on the combined Yale B database show that the proposed method can extract more robust illumination invariants, which improves face recognition rate.

As seen in stream data processing, it is necessary to extract a particular data field from bulk data, where we can use a field-data extractor. Particularly, an (M,N)-field-data extractor reads out any consecutive N bytes from an M-byte register by connecting its input/output using multiplexers (MUXs). However, the number of required MUXs increases too much as the input/output byte widths increase. It is known that partitioning a MUX network leads to reducing the number of MUXs. In this paper, we firstly pick up a multi-layered MUX network, which is generated by repeatedly partitioning a MUX network into a collection of single-layered MUX networks. We show that the multi-layered MUX network is equivalent to the barrel shifter from which redundant MUXs and wires are removed, and we prove that the number of required MUXs becomes the smallest among MUX-network-partitioning based field-data extractors. Next, we propose a rotator-based MUX network for a field-data extractor, which is based on reading out a particular data in an input register to a rotator. The byte width of the rotator is the same as its output register and hence we no longer require any extra wires nor MUXs. By rotating the input data appropriately, we can finally have a right-ordered data into an output register. Experimental results show that a multi-layered MUX network reduces the number of required gates to construct a field-data extractor by up to 97.0% compared with the one using a naive approach and its delay becomes 1.8ns-2.3ns. A rotator-based MUX network with a control circuit also reduces the number of required gates to construct a field-data extractor by up to 97.3% compared with the one using a naive approach and its delay becomes 2.1ns-2.9ns.

This paper proposes a novel multiple chaos embedded gravitational search algorithm (MCGSA) that simultaneously utilizes multiple different chaotic maps with a manner of local search. The embedded chaotic local search can exploit a small region to refine solutions obtained by the canonical gravitational search algorithm (GSA) due to its inherent local exploitation ability. Meanwhile it also has a chance to explore a huge search space by taking advantages of the ergodicity of chaos. To fully utilize the dynamic properties of chaos, we propose three kinds of embedding strategies. The multiple chaotic maps are randomly, parallelly, or memory-selectively incorporated into GSA, respectively. To evaluate the effectiveness and efficiency of the proposed MCGSA, we compare it with GSA and twelve variants of chaotic GSA which use only a certain chaotic map on a set of 48 benchmark optimization functions. Experimental results show that MCGSA performs better than its competitors in terms of convergence speed and solution accuracy. In addition, statistical analysis based on Friedman test indicates that the parallelly embedding strategy is the most effective for improving the performance of GSA.

In this paper, we examine the locality property of the original Fractional Repetition (FR) codes and propose two constructions for FR codes with better locality. For this, we first derive the capacity of the FR codes with locality 2, that is the maximum size of the file that can be stored. Construction 1 generates an FR code with repetition degree 2 and locality 2. This code is optimal in the sense of achieving the capacity we derived. Construction 2 generates an FR code with repetition degree 3 and locality 2 based on 4-regular graphs with girth g. This code is also optimal in the same sense.

When haze exists in an image of an outdoor scene, the visibility of objects in the image is deteriorated. In recent years, to improve the visibility of objects in such images, many dehazing methods have been investigated. Most of the methods are based on the atmospheric scattering model. In such methods, the transmittance and global atmospheric light are estimated from an input image and a dehazed image is obtained by substituting them into the model. To estimate the transmittance and global atmospheric light, the dark channel prior is a major and powerful concept that is employed in many dehazing methods. In this paper, we propose a new dehazing method in which the degree of haze removal can be adjusted by changing its parameters. Our method is also based on the atmospheric scattering model and employs the dark channel prior. In our method, the estimated transmittance is adjusted to a more suitable value by a transform function. By choosing appropriate parameter values for each input image, good haze removal results can be obtained by our method. In addition, a quantitative index for evaluating the quality of a dehazed image is proposed in this paper. It can be considered that haze removal is a type of saturation enhancement. On the other hand, an output image obtained using the atmospheric scattering model is generally darker than the input image. Therefore, we evaluate the quality of dehazed images by considering the balance between the brightness and saturation of the input and output images. The validity of the proposed index is examined using our dehazing method. Then a comparison between several dehazing methods is carried out using the index. Through these experiments, the effectiveness of our dehazing method and the quantitative index is confirmed.

The purpose of this paper is to propose a flexible load-dependent digital soft-start control method for dc-dc converters in a 380Vdc system. The soft-start operation is needed to prevent negative effects such as large inrush current and output overshoot to a power supply in the start-up process of dc-dc converters. In the conventional soft-start operation, a dc-dc converter has a very slow start-up to deal with the light load condition. Therefore, it always takes a long time in any load condition to start up a power supply and obtain the desired output. In the proposed soft-start control method, the speed of the start-up process is flexibly controlled depending on the load condition. To obtain the optimal speed for any load condition, the speed of the soft-start is determined from a approximated function of load current, which is estimated from experiment results in advance. The proposed soft-start control method is evaluated both in simulations and experiments. From results, it is confirmed that the proposed method has superior soft-start characteristics compared to the conventional one.

Due to the rapid growth of applications that are based on Internet of Things (IoT) and real-time communications, mobile traffic growth is increasing exponentially. In highly populated areas, sudden concentration of numerous mobile user traffic can cause radio resource shortage, where traffic offloading is essential in preventing overload problems. Vertical handover (VHO) technology which supports seamless connectivity across heterogeneous wireless networks is a core technology of traffic offloading. In VHO, minimizing service interruption is a key design factor, since service interruption deteriorates service performance and degrades user experience (UX). Although 3GPP standard VHO procedures are designed to prevent service interruption, severe quality of service (QoS) degradation and severe interruption can occur in real network environments due to unintended disconnections with one's base station (BS) or access point (AP). In this article, the average minimum handover interruption time (HIT) (i.e., the guaranteed HIT influence) between LTE and Wi-Fi VHO is analyzed and measured based on 3GPP VHO access and decision procedures. In addition, the key parameters and procedures which affect HIT performance are analyzed, and a reference probability density function (PDF) for HIT prediction is derived from Kolmogorov-Smirnov test techniques.

This paper describes a logarithmic compression ADC using a subranging TDC and the transient response of a comparator. We utilized the settling time of the comparator for a logarithmic compression instead of a logarithmic amplifier. The settling time of the comparator is inversely proportional to the logarithm of an input voltage. In the proposed ADC, an input voltage is converted into a pulse whose width represents the settling time of the comparator. Subsequently, the TDC converts the pulse width into a binary code. The supply voltage of the proposed ADC can be reduced more than a conventional logarithmic ADC because an analog to digital conversion takes place in the time domain. We confirmed through a 0.18-µm CMOS circuit simulation that the proposed ADC achieves a resolution of 11 bits, a sampling rate of 20 MS/s, a dynamic range of 59 dB and a power consumption of 9.8 mW at 1.5 V operation.

Particle swarm optimizer network (PSON) is one of the multi-swarm PSOs. In PSON, a population is divided into multiple sub-PSOs, each of which searches a solution space independently. Although PSON has a good solving performance, it may be trapped into a local optimum solution. In this paper, we introduce into PSON a dynamic stochastic network topology called “PSON with stochastic connection” (PSON-SC). In PSON-SC, each sub-PSO can be connected to the global best (gbest) information memory and refer to gbest stochastically. We show clearly herein that the diversity of PSON-SC is higher than that of PSON, while confirming the effectiveness of PSON-SC by many numerical simulations.

In this paper, we investigate a user and antenna joint selection problem in multi-user large-scale MIMO downlink networks, where a BS with N transmit antennas serves K users, and N is much larger than K. The BS activates only S(S≤N) antennas for data transmission to reduce hardware cost and computation complexity, and selects the set of users to which data is to be transmitted by maximizing the sum-rate. The optimal user and antenna joint selection scheme based on exhaustive search causes considerable computation complexity. Thus, we propose a new joint selection algorithm with low complexity and analyze the performance of the proposed scheme in terms of sum-rate and complexity. When S=7, N=10, K=5, and SNR=10dB, the sum-rate of the proposed scheme is 5.1% lower than that of the optimal scheme, while the computation complexity of the proposed scheme is reduced by 99.0% compared to that of the optimal scheme.

In this letter, we focus on the subcarrier allocation problem for device-to-device (D2D) communication in cellular networks to improve the cellular energy efficiency (EE). Our goal is to maximize the weighted cellular EE and its solution is obtained by using a game-theoretic learning approach. Specifically, we propose a lower bound instead of the original optimization objective on the basis of the proven property that the gap goes to zero as the number of transmitting antennas increases. Moreover, we prove that an exact potential game applies to the subcarrier allocation problem and it exists the best Nash equilibrium (NE) which is the optimal solution to optimize the lower bound. To find the best NE point, a distributed learning algorithm is proposed and then is proved that it can converge to the best NE. Finally, numerical results verify the effectiveness of the proposed scheme.

In recent years, covariance descriptors have received considerable attention as a strong representation of a set of points. In this research, we propose a new metric learning algorithm for covariance descriptors based on the Dykstra algorithm, in which the current solution is projected onto a half-space at each iteration, and which runs in O(n3) time. We empirically demonstrate that randomizing the order of half-spaces in the proposed Dykstra-based algorithm significantly accelerates convergence to the optimal solution. Furthermore, we show that the proposed approach yields promising experimental results for pattern recognition tasks.

The quality of codebook is very important in visual image classification. In order to boost the classification performance, a scheme of codebook generation for scene image recognition based on parallel key SIFT analysis (PKSA) is presented in this paper. The method iteratively applies classical k-means clustering algorithm and similarity analysis to evaluate key SIFT descriptors (KSDs) from the input images, and generates the codebook by a relaxed k-means algorithm according to the set of KSDs. With the purpose of evaluating the performance of the PKSA scheme, the image feature vector is calculated by sparse code with Spatial Pyramid Matching (ScSPM) after the codebook is constructed. The PKSA-based ScSPM method is tested and compared on three public scene image datasets. The experimental results show the proposed scheme of PKSA can significantly save computational time and enhance categorization rate.

Personalized PageRank (PPR) is a typical similarity metric between nodes in a graph, and node searches based on PPR are widely used. In many applications, graphs change dynamically, and in such cases, it is desirable to perform ad hoc searches based on PPR. An ad hoc search involves performing searches by varying the search parameters or graphs. However, as the size of a graph increases, the computation cost of performing an ad hoc search can become excessive. In this paper, we propose a method called Castanet that offers fast ad hoc searches of PPR. The proposed method features (1) iterative estimation of the upper and lower bounds of PPR scores, and (2) dynamic pruning of nodes that are not needed to obtain a search result. Experiments confirm that the proposed method does offer faster ad hoc PPR searches than existing methods.

Peer-to-peer (P2P) overlay networks are widely employed in distributed systems. The number of hops required by a node to locate an object is the fundamental search cost of a P2P network. Creating replicas can efficiently reduce the cost of object search, so how to deploy replicas to reduce the cost as much as possible is a critical problem of P2P networks. In the literature, most existing replica placement strategies arrange replicas at nodes near the one containing the considered object. In this paper, we formally demonstrate that for a complete Chord P2P network and many non-complete Chord ones, due to their deterministic structures, we can allocate replicas to nodes closest to the target in the identifier space to maximize the reduction in the total number of hops required by all nodes to reach a copy of the object during the search heading to the target node.

In this paper, we propose the application of principal component analysis (PCA) to scale-spaces. PCA is a standard method used in computer vision. Because the translation of an input image into scale-space is a continuous operation, it requires the extension of conventional finite matrix-based PCA to an infinite number of dimensions. Here, we use spectral theory to resolve this infinite eigenvalue problem through the use of integration, and we propose an approximate solution based on polynomial equations. In order to clarify its eigensolutions, we apply spectral decomposition to Gaussian scale-space and scale-normalized Laplacian of Gaussian (sLoG) space. As an application of this proposed method, we introduce a method for generating Gaussian blur images and sLoG images, demonstrating that the accuracy of such an image can be made very high by using an arbitrary scale calculated through simple linear combination. Furthermore, to make the scale-space filtering efficient, we approximate the basis filter set using Gaussian lobes approximation and we can obtain XY-Separable filters. As a more practical example, we propose a new Scale Invariant Feature Transform (SIFT) detector.

PPDP (Privacy-Preserving Data Publishing) is technology that discloses personal information while protecting individual privacy. k-anonymity is a privacy model that should be achieved in PPDP. However, k-anonymity does not guarantee privacy against adversaries who have knowledge of even a few uncommon individuals in a population. In this paper, we propose a new model, called k-presence-secrecy, that prevents such adversaries from inferring whether an arbitrary individual is included in a personal data table. We also propose an algorithm that satisfies the model. k-presence-secrecy is a practical model because an algorithm that satisfies it requires only a PPDP target table as personal information, whereas previous models require a PPDP target table and almost all the background knowledge of adversaries. Our experiments show that, whereas an algorithm satisfying only k-anonymity cannot protect privacy, even against adversaries who have knowledge for one uncommon individual in a population, our algorithm can do so with less information loss and shorter execution time.

Distributed compressive video sensing (DCVS) is a new paradigm for low-complexity video compression. To achieve the highest possible perceptual coding performance under the measurements budget constraint, we propose a perceptual optimized DCVS codec by jointly exploiting the reweighted sampling and rate-distortion optimized measurements allocation technologies. A visual saliency modulated just-noticeable distortion (VS-JND) profile is first developed based on the side information (SI) at the decoder side. Then the estimated correlation noise (CN) between each non-key frame and its SI is suppressed by the VS-JND. Subsequently, the suppressed CN is utilized to determine the weighting matrix for the reweighted sampling as well as to design a perceptual rate-distortion optimization model to calculate the optimal measurements allocation for each non-key frame. Experimental results indicate that the proposed DCVS codec outperforms the other existing DCVS codecs in term of both the objective and subjective performance.

Video quality generally suffers from packet losses caused by an unreliable channel when video is transmitted over an error-prone wireless channel. This quality degradation is the main reason that a video compression encoder uses error-resilient coding to deal with the high packet-loss probability. The use of adequate error resilience can mitigate the effects of channel errors, but the coding efficiency for bit reduction will be decreased. On the other hand, H.264/AVC uses multiple reference frame (MRF) motion compensation for a higher coding efficiency. However, an increase in the number of reference frames in the H.264/AVC encoder has been recently observed, making the received video quality worse in the presence of transmission errors if the cyclic intra-refresh is used as the error-resilience method. This is because the reference-block selection in the MRF chooses blocks on the basis of the rate distortion optimization, irrespective of the intra-refresh coding. In this paper, a new error-resilient reference selection method is proposed to provide error resilience for MRF based motion compensation. The proposed error-resilient reference selection method achieves an average PSNR enhancement up to 0.5 to 2dB in 10% packet-loss-ratio environments. Therefore, the proposed method can be valuable in most MRF-based interactive video encoding system, which can be used for video broadcasting and mobile video conferencing over an erroneous network.

To analyze large-scale data efficiently, studies on Hadoop, one of the most popular MapReduce frameworks, have been actively done. Meanwhile, most of the large-scale data analysis applications, e.g., data clustering, are required to do the same map and reduce functions repeatedly. However, Hadoop cannot provide an optimal performance for iterative MapReduce jobs because it derives a result by doing one phase of map and reduce functions. To solve the problems, in this paper, we propose a new efficient resource management framework for iterative MapReduce processing in large-scale data analysis. For this, we first design an iterative job state-machine for managing the iterative MapReduce jobs. Secondly, we propose an invariant data caching mechanism for reducing the I/O costs of data accesses. Thirdly, we propose an iterative resource management technique for efficiently managing the resources of a Hadoop cluster. Fourthly, we devise a stop condition check mechanism for preventing unnecessary computation. Finally, we show the performance superiority of the proposed framework by comparing it with the existing frameworks.