In this paper, we present a hybrid speech emotion recognition system exploiting both spectral and prosodic features in speech. For capturing the emotional information in the spectral domain, we propose a new spectral feature extraction method by applying a novel non-uniform subband processing, instead of the mel-frequency subbands used in Mel-Frequency Cepstral Coefficients (MFCC). For prosodic features, a set of features that are closely correlated with speech emotional states are selected. In the proposed hybrid emotion recognition system, due to the inherently different characteristics of these two kinds of features (e.g., data size), the newly extracted spectral features are modeled by Gaussian Mixture Model (GMM) and the selected prosodic features are modeled by Support Vector Machine (SVM). The final result of the proposed emotion recognition system is obtained by combining the results from these two subsystems. Experimental results show that (1) the proposed non-uniform spectral features are more effective than the traditional MFCC features for emotion recognition; (2) the proposed hybrid emotion recognition system using both spectral and prosodic features yields the relative recognition error reduction rate of 17.0% over the traditional recognition systems using only the spectral features, and 62.3% over those using only the prosodic features.

A characteristic-based constrained interpolation profile (CIP) method for solving three-dimensional, time-dependent Maxwell's equations is successfully developed. It is utilized to solve one-dimensional wave equations in the formulation of the Maxwell's equations. Calculation procedure of the CIP method for three-dimensional scattering analysis is described in details. Update equations for boundary conditions of a perfectly conducting (PEC) interface and a dielectric interface are formulated and obtained in explicit forms. Numerical analyses of electromagnetic scatterings of PEC sphere, dielectric sphere and PEC cube are performed and the scattering coefficient is calculated and compared with the Mie's analytic results. As a result, the scattering coefficients show good agreement with the Mie's results, which demonstrates the validity of the CIP method and the formulated update equations. It is also shown that the phase of the scattering coefficients determined by the CIP method are slightly more accurate than that of the FDTD method.

Simple power analysis (SPA) can be employed in examining the power consumption trace of elliptic curve scalar multiplication to retrieve the computational sequence. However, SPA cannot distinguish point addition from point subtraction. The attacker still requires an exhaustive search to recover the private key when it is recoded in NAF or recoded by the 2-bit sliding window method. The average Hamming weight of an n-bit NAF recoded scalar is n/3, and an exhaustive search among the 2n/3 candidates is required. This paper shows that in a left-to-right NAF recoded or a left-to-right 2-bit sliding window manipulated scalar the relative position of nonzero bits will reveal their values. Our analysis skill reduces the number of candidates of the scalar from the naive search of 2n/3 to 22n/9 and 20.19n respectively for the cases of NAF and sliding window method.

Narrow-wall-connected microstrip-to-waveguide transition using V-shaped patch element in millimeter-wave band is proposed. Since the microstrip line on the narrow-wall is perpendicular to the E-plane of the waveguide, waveguide field does not couple directly to the microstrip line. The current on the V-shaped patch element flows along inclined edges, then current on the V-shaped patch element couples to the microstrip line efficiently. Three types of transitions are investigated. A numerical investigation of these transitions show some relations between bandwidth and insertion loss. It is confirmed that the improved transition exhibits an insertion loss of 0.6 dB from 76 to 77 GHz, and a bandwidth of 4.1% (3.15 GHz) for the reflection coefficient below -15 dB.

A waveform of an ultra wideband impulse radio (UWB-IR) system can be extremely distorted through a channel even for free-space transmission because of antenna dispersion. This highly degrades the link budget performance. Therefore, the understand of antenna characteristics, which effects on waveform distortion, is necessary. This paper studies the waveform distortion due to antenna in free space transmission in UWB-IR system. The link budget is usually evaluated by using the Friis' transmission formula. However, it is not directly applicable to the UWB-IR transmission system. The link budget evaluation formula attended from conventional Friis' transmission formula that takes into account the transmitted waveform, its distortion due to the antennas, the channel and the correlation receiver is proposed. Since the antenna is significant pulse-shaping filters in UWB-IR system, the example kind of the log-periodic dipole antenna (LPDA) is experimentally examined, especially focused on the effect of the template waveforms.

This paper presents an approach for improving proximity and diversity in multiobjective evolutionary algorithms (MOEAs). The idea is to discover new nondominated solutions in the promising area of search space. It can be achieved by applying mutation only to the most converged and the least crowded individuals. In other words, the proximity and diversity can be improved because new nondominated solutions are found in the vicinity of the individuals highly converged and less crowded. Empirical results on multiobjective knapsack problems (MKPs) demonstrate that the proposed approach discovers a set of nondominated solutions much closer to the global Pareto front while maintaining a better distribution of the solutions.

Design of high gain and high efficiency antennas is one of the key challenges in antenna engineering and especially in millimeter wave communication systems. Various types of planar waveguide arrays with series-fed traveling wave operation have been developed in Tokyo Tech with the special focus upon efficiency enhancement as well as reduction of fabrication cost. In this review, four kinds of single layer waveguide arrays characterized with the series fed travelling wave operation are surveyed first. To cope with the bandwidth narrowing effects due to long line effects associated with the series fed operation, authors have introduced partially corporate feed embedded in the single layer waveguide. They further extended the study to cover fully corporate feed arrays with multiple layer waveguide as well; a new fabrication technique of diffusion bonding of laminated thin plates has the potential to realize the low cost mass production of multi-layer structures for the millimeter wave application. Secondly, the novel methods for loss evaluation of copper plate substrate are established for the design of post-wall waveguide arrays where dielectric loss and conductor loss is determined in wide range of millimeter wave band, by using the Whispering gallery mode resonator. This enables us to design the planar arrays with the loss taken into account. Finally, the planar arrays are now applied to two kinds of systems in the Tokyo Tech millimeter wave project; the indoor short range file-transfer systems and the outdoor communication systems for the medium range backhaul links. The latter has been field-tested in the model network built in Tokyo Tech Ookayama campus. Early stage progress of the project including unique propagation data is also reported.

We address the problem of perceived age estimation from face images, and propose a new semi-supervised approach involving two novel aspects. The first novelty is an efficient active learning strategy for reducing the cost of labeling face samples. Given a large number of unlabeled face samples, we reveal the cluster structure of the data and propose to label cluster-representative samples for covering as many clusters as possible. This simple sampling strategy allows us to boost the performance of a manifold-based semi-supervised learning method only with a relatively small number of labeled samples. The second contribution is to take the heterogeneous characteristics of human age perception into account. It is rare to misjudge the age of a 5-year-old child as 15 years old, but the age of a 35-year-old person is often misjudged as 45 years old. Thus, magnitude of the error is different depending on subjects' age. We carried out a large-scale questionnaire survey for quantifying human age perception characteristics, and propose to utilize the quantified characteristics in the framework of weighted regression. Consequently, our proposed method is expressed in the form of weighted least-squares with a manifold regularizer, which is scalable to massive datasets. Through real-world age estimation experiments, we demonstrate the usefulness of the proposed method.

Fuzzy cognitive maps (FCMs) are used to support decision-making, and the decision processes are performed by inference of FCMs. The inference greatly depends on activation functions such as sigmoid function, hyperbolic tangent function, step function, and threshold linear function. However, the sigmoid functions widely used for decision-making processes have been designed by experts. Therefore, we propose a method for designing sigmoid functions through Lyapunov stability analysis. We show the usefulness of the proposed method through the experimental results in inference of FCMs using the designed sigmoid functions.

Estimating the number of signals (NIS) is an important goal in array signal processing, such as direction-of-arrival (DOA) estimation. A common approach for solving this problem is to use an eigenvalue of the array covariance matrix and information criterion, such as the Akaike information criterion (AIC) and minimum description length (MDL). However they suffer serious degradation, when the incoming signals are coherent. To estimate the NIS of the coherent signals impinging on a uniform linear array (ULA), a method for estimating the number of signals without eigendecomposition (MENSE) is proposed. The accuracy of the NIS estimation performance of MENSE is superior to the other algorithms equipped with preprocessing such as the spatial smoothing preprocessing (SSP) and forward/backward spatial smoothing techniques (FBSS) to decorrelate the coherency of signals. Instead of using SSP or FBSS preprocessing, MENSE uses the Hankel correlation matrices. The Hankel correlation matrices can not only decorrelate the coherency of signals but also suppress the influence of noise. However, in severe conditions like low signal-to-noise ratio (SNR) or a closely spaced signals impinging on a ULA, the NIS estimation metric of MENSE has some bias which causes estimation error. In this paper, we pay attention to the multiplicity defined by the ratio of the geometric mean to the arithmetic mean. Accordingly, we propose a new estimation metric that has less bias than that in MENSE. The Computer simulation results show that the proposed method is superior to MENSE in the above severe conditions.

Delay Tolerant Networks (DTNs) are a class of emerging networks that experience frequent and long-duration partitions. Delay is inevitable in DTNs, so ensuring the validity and reliability of the message transmission and making better use of buffer space are more important than concentrating on how to decrease the delay. In this paper, we present a novel routing protocol named Location and Direction Aware Priority Routing (LDPR) for DTNs, which utilizes the location and moving direction of nodes to deliver a message from source to destination. A node can get its location and moving direction information by receiving beacon packets periodically from anchor nodes and referring to received signal strength indicator (RSSI) for the beacon. LDPR contains two schemes named transmission scheme and drop scheme, which take advantage of the nodes' information of the location and moving direction to transmit the message and store the message into buffer space, respectively. Each message, in addition, is branded a certain priority according to the message's attributes (e.g. importance, validity, security and so on). The message priority decides the transmission order when delivering the message and the dropping sequence when the buffer is full. Simulation results show that the proposed LDPR protocol outperforms epidemic routing (EPI) protocol, prioritized epidemic routing (PREP) protocol, and DTN hierarchical routing (DHR) protocol in terms of packet delivery ratio, normalized routing overhead and average end-to-end delay. It is worth noting that LDPR doesn't need infinite buffer size to ensure the packet delivery ratio as in EPI. In particular, even though the buffer size is only 50, the packet delivery ratio of LDPR can still reach 93.9%, which can satisfy general communication demand. We expect LDPR to be of greater value than other existing solutions in highly disconnected and mobile networks.

The security of permutation-based hash functions in the ideal permutation model has been studied when the input-length of compression function is larger than the input-length of the permutation function. In this paper, we consider permutation based compression functions that have input lengths shorter than that of the permutation. Under this assumption, we propose a permutation based compression function and prove its security with respect to collision and (second) preimage attacks in the ideal permutation model. The proposed compression function can be seen as a generalization of the compression function of MD6 hash function.

A new linearly tapered slot antenna (LTSA) with defected sides is proposed in this letter. Both sides are defected with half-dumbbell shape slots that may alter the surface current intensities on both sides. As the half-dumbbell size is increased, the 3-dB beamwidth of the proposed antenna is 4° and 6° lower in the E/H-plane, respectively, than these of the LTSA without defects. Accordingly, the measured gain is improved by up to 3.75 dB and the first side lobe level is lowered by about -10.8 dB and -5.8 dB in the E/H-planes, respectively.

Broadband antennas have various applications in digital terrestrial television (DTV) services. Compact broadband antennas are required for arranging in long and narrow space along the rim of a laptop display. A leaky-wave antenna using the composite right/left-handed transmission line (CRLH-TL) is one of the candidates for achieving the broadband antenna. However, there are not enough to design guideline of small leaky wave antennas using the CRLH-TL for UHF band. In this paper, a CRLH-TL comprising a ladder network is proposed for broadband and simple structure. The paper also discusses the design of a leaky-wave antenna with the CRLH-TL operating in the DTV band. The relation between the operating bandwidth and attenuation constant of the CRLH-TL is discussed. An antenna that can be accommodated in the limited and narrow space available in mobile terminals has to be designed. Hence, the effects of the number of cells and a finite ground plane are discussed with the purpose of achieving the miniaturization of the antenna. In this study, the transmission and radiation characteristics of the fabricated antennas are measured. The gain of the fabricated antenna is confirmed to remain almost constant even when the operating frequency is varied. The maximum gain and operating band achieved in this study are approximately -0.6 dBi and about 54%, respectively.

Regression analysis that incorporates measurement errors in input variables is important in various applications. In this study, we consider this problem within a framework of Gaussian process regression. The proposed method can also be regarded as a generalization of kernel regression to include errors in regressors. A Markov chain Monte Carlo method is introduced, where the infinite-dimensionality of Gaussian process is dealt with a trick to exchange the order of sampling of the latent variable and the function. The proposed method is tested with artificial data.

Kernel logistic regression (KLR) is a powerful and flexible classification algorithm, which possesses an ability to provide the confidence of class prediction. However, its training--typically carried out by (quasi-)Newton methods--is rather time-consuming. In this paper, we propose an alternative probabilistic classification algorithm called Least-Squares Probabilistic Classifier (LSPC). KLR models the class-posterior probability by the log-linear combination of kernel functions and its parameters are learned by (regularized) maximum likelihood. In contrast, LSPC employs the linear combination of kernel functions and its parameters are learned by regularized least-squares fitting of the true class-posterior probability. Thanks to this linear regularized least-squares formulation, the solution of LSPC can be computed analytically just by solving a regularized system of linear equations in a class-wise manner. Thus LSPC is computationally very efficient and numerically stable. Through experiments, we show that the computation time of LSPC is faster than that of KLR by two orders of magnitude, with comparable classification accuracy.

A C-testable 4-2 adder tree for an easily testable high-speed multiplier is proposed, and a recursive method for test generation is shown. By using the specific patterns that we call 'alternately inverted patterns,' the adder tree, as well as partial product generators, can be tested with 14 patterns regardless of its operand size under the cell fault model. The test patterns are easily fed through the partial product generators. The hardware overhead of the 4-2 adder tree with partial product generators for a 64-bit multiplier is about 15%. By using a previously proposed easily testable adder as the final adder, we can obtain an easily testable high-speed multiplier.

A new scheme to avoid null zone for HF-band RFID without expanding antenna size is proposed. At first, we demonstrate by FDTD simulation that the null zone occurs because of cancellation of magnetic fields over the loop surface. To prevent cancellation of magnetic fields, the loop antenna is split into four parts, which work as a planar array antenna. The outputs of antennas are gathered by using combining circuit. We have validated by FDTD simulation that the proposed scheme enlarges the worst received power by 13.1 dB.

This paper proposes a channel capacity maximization method for Multiple-Input Multiple-Output (MIMO) antennas with parasitic elements. Reactive terminations are connected to the parasitic elements, and the reactance values are determined to achieve stochastically high channel capacity for the environment targeted. This method treats the S-parameter and propagation channel of the antenna, including the parasitic elements, as a combined circuit. The idea of the 'parasitic channel,' which is observed at the parasitic antenna, is introduced to simplify the optimization procedure. This method can significantly reduce the number of necessary measurements of the channel for designing the antenna. As a design example, a bidirectional Yagi-Uda array, which has two driven antennas at both ends of the linear array, is measured in an indoor environment. The resulting design offers enhanced channel capacity mainly due to its improved signal-to-noise ratio compared to the antenna without the parasitic antennas.

This paper presents a combined feature extraction method to improve the performance of bag-of-features image classification. We apply 10 relevant operations to global/local statistics of visual words. Because the pairwise combination of visual words is large, we apply feature selection methods including fisher discriminant criterion and L1-SVM. The effectiveness of the proposed method is confirmed through the experiment.