Minook KIM Tae-Jun LEE Hyung-Min PARK
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.
When there are multiple component predictors, it is promising to integrate them into one predictor for advanced reasoning. If each component predictor is given as a stochastic model in the form of probability distribution, an exponential mixture of the component probability distributions provides a good way to integrate them. However, weight parameters used in the exponential mixture model are difficult to estimate if there is no training samples for performance evaluation. As a suboptimal way to solve this problem, weight parameters may be estimated so that the exponential mixture model should be a balance point that is defined as an equilibrium point with respect to the distance from/to all component probability distributions. In this paper, we propose a weight parameter estimation method that represents this concept using a symmetric Kullback-Leibler divergence and generalize this method.
Template tracking has been extensively studied in Computer Vision with a wide range of applications. A general framework is to construct a parametric model to predict movement and to track the target. The difference in intensity between the pixels belonging to the current region and the pixels of the selected target allows a straightforward prediction of the region position in the current image. Traditional methods track the object based on the assumption that the relationship between the intensity difference and the region position is linear or non-linear. They will result in bad tracking performance when just one model is adopted. This paper proposes a method, called as Mixture Hyperplanes Approximation, which is based on finite mixture of generalized linear regression models to perform robust tracking. Moreover, a fast learning strategy is discussed, which improves the robustness against noise. Experiments demonstrate the performance and stability of Mixture Hyperplanes Approximation.
Peng SONG Wenming ZHENG Xinran ZHANG Yun JIN Cheng ZHA Minghai XIN
Most of the current voice conversion methods are conducted based on parallel speech, which is not easily obtained in practice. In this letter, a novel iterative speaker model alignment (ISMA) method is proposed to address this problem. First, the source and target speaker models are each trained from the background model by adopting maximum a posteriori (MAP) algorithm. Then, a novel ISMA method is presented for alignment and transformation of spectral features. Finally, the proposed ISMA approach is further combined with a Gaussian mixture model (GMM) to improve the conversion performance. A series of objective and subjective experiments are carried out on CMU ARCTIC dataset, and the results demonstrate that the proposed method significantly outperforms the state-of-the-art approach.
Irneza ISMAIL Quang NGUYEN-THE Motoharu MATSUURA Naoto KISHI
All-optical 1-to-6 wavelength multicasting of a 10-Gb/s picosecond-tunable-width converted return-to-zero (RZ)-on-off-keying (OOK) data signal using a wideband-parametric pulse source from a distributed Raman amplifier (DRA) is experimentally demonstrated. Width-tunable wavelength multicasting within the C-band with approximately 40.6-nm of separation with various compressed RZ data signal inputs have been proposed and demonstrated. The converted multicast pulse widths can be flexibly controlled down to 2.67 ps by tuning the Raman pump powers of the DRA. Nearly equal pulse widths at all multicast wavelengths are obtained. Furthermore, wide open eye patterns and penalties less than 1.2 dB at the 10-9 bit-error-rate (BER) level are found.
Quynh NGUYEN QUANG NHU Quang NGUYEN-THE Hung NGUYEN TAN Motoharu MATSUURA Naoto KISHI
A combination of nonreturn-to-zero (NRZ)-to-return-to-zero (RZ) waveform conversion and wavelength multicasting with pulsewidth tunability is experimentally demonstrated. A NRZ data signal is injected into a highly nonlinear fiber (HNLF)-based four-wave mixing (FWM) switch with four RZ clocks compressed by a Raman amplification-based multiwavelength pulse compressor (RA-MPC). The NRZ signal is multicast and converted to RZ signals in a continuously wide pulsewidth tuning range between around 12.17 and 4.68 ps by changing the Raman pump power of the RA-MPC. Error-free operations of the converted RZ signals with different pulsewidths are achieved with negative power penalties compared with the back-to-back NRZ signal and the small variation among received powers of RZ output channels at a bit-error-rate (BER) of 10-9. The NRZ-to-RZ waveform conversion and wavelength multicasting without using the RA-MPC are also successfully implemented.
Daliso BANDA Mizuki MOTOYOSHI Tomokazu KOIZUMI Osamu WADA Tuan Thanh TA Suguru KAMEDA Noriharu SUEMATSU Tadashi TAKAGI Kazuo TSUBOUCHI
RF under sampling is more suitable for Satellite receiver systems in comparison to terrestrial systems. In conventional RF under sampling the minimum sampling frequency (fs) should be atleast twice the system bandwidth; therefore for a system with a wide bandwidth, a relatively high fs is necessary. In this paper we propose a direct RF under sampling reception method that halves fs. The proposed f's is achieved by folding in band noise in half. A method of adapting f's for the reception of signals in different channels is also proposed; this ensures that the SNR is not degraded for any channel. To evaluate the proposed technique's performance and compare it to the conventional case a 3 channel, 1 GHz band test receiver and it's key device (i.e. S/H circuit) are developed. Using SNR and EVM as performance indexes, the performance of the proposed technique has been evaluated and compared to that of the conventional technique. The evaluation results show that the proposed technique can achieve the same performance as conventional RF under sampling for all 3 channels, using only half of the sampling frequency of the conventional technique.
Takana KAHO Yo YAMAGUCHI Hiroyuki SHIBA Tadao NAKAGAWA Kazuhiro UEHARA Kiyomichi ARAKI
Novel multi-band mixers that can receive multiple band signals concurrently are proposed and evaluated. The mixers achieve independent gain control through novel relative power control method of the multiple local oscillator (LO) signals. Linear control is also achieved through multiple LO signal input with total LO power control. Theoretical analysis shows that odd-order nonlinearity components of the multiple LO signals support linear conversion gain control. Dual- and triple-band tests are conducted using typical three MOSFET mixers fabricated by a 0.25 µm SiGe BiCMOS process. Measurements confirm over 40 dB independent control of conversion gain, linear control achieved through LO input power control. The proposed mixers have high input linearity with a 5 dBm output third intercept point. A method is also proposed to reduce interference caused by mixing between multiple LO signals.
Kosei KURISU Nobuo SUEMATSU Kazunori IWATA Akira HAYASHI
In image segmentation, finite mixture modeling has been widely used. In its simplest form, the spatial correlation among neighboring pixels is not taken into account, and its segmentation results can be largely deteriorated by noise in images. We propose a spatially correlated mixture model in which the mixing proportions of finite mixture models are governed by a set of underlying functions defined on the image space. The spatial correlation among pixels is introduced by putting a Gaussian process prior on the underlying functions. We can set the spatial correlation rather directly and flexibly by choosing the covariance function of the Gaussian process prior. The effectiveness of our model is demonstrated by experiments with synthetic and real images.
Ikuma ANDO Gia Khanh TRAN Kiyomichi ARAKI Takayuki YAMADA Takana KAHO Yo YAMAGUCHI Tadao NAKAGAWA
In this paper we describe and experimentally validate a dual-band digital predistortion (DPD) model we propose that takes account of the intermodulation and harmonic distortion produced when the center frequencies of input bands have a harmonic relationship. We also describe and experimentally validate our proposed novel dual-band power amplifier (PA) linearization architecture consisting of a single feedback loop employing a dual-band mixer. Experiment results show that the DPD linearization the proposed model provides can compensate for intermodulation and harmonic distortion in a way that the conventional two-dimensional (2-D) DPD approach cannot. The proposed feedback architecture should make it possible to simplify analog-to-digital converter (ADC) design and eliminate the time lag between different feedback paths.
Recently, a next-generation heterodyne mixer detector---a hot electron bolometer (HEB) mixer employing a superconducting microbridge---has gradually opened up terahertz-band astronomy. The surrounding state-of-the-art technologies including fabrication processes, 4 K cryostats, cryogenic low-noise amplifiers, local oscillator sources, micromachining techniques, and spectrometers, as well as the HEB mixers, have played a valuable role in the development of super-low-noise heterodyne spectroscopy systems for the terahertz band. The current developmental status of terahertz-band HEB mixer receivers and their applications for spectroscopy and astronomy with ground-based, airborne, and satellite telescopes are presented.
Koichi KOBAYASHI Takuro NAGAMI Kunihiko HIRAISHI
In this paper, optimal control of multi-vehicle systems is studied. In the case where collision avoidance between vehicles and obstacle avoidance are imposed, state discretization is effective as one of the simplified approaches. Furthermore, using state discretization, cooperative actions such as rendezvous can be easily specified by linear temporal logic (LTL) formulas. However, it is not necessary to discretize all states, and partial states (e.g., the position of vehicles) should be discretized. From this viewpoint, a new control method for multi-vehicle systems is proposed in this paper. First, the system in which partial states are discretized is formulated. Next, the optimal control problem with constraints described by LTL formulas is formulated, and its solution method is proposed. Finally, numerical simulations are presented. The proposed method provides us a useful method in control of multi-vehicle systems.
Hanchao ZHOU Ning ZHU Wei LI Zibo ZHOU Ning LI Junyan REN
A monolithic frequency synthesizer with wide tuning range, low phase noise and spurs was realized in 0.13,$mu$m CMOS technology. It consists of an analog PLL, a harmonic-rejection mixer and injection-locked frequency doublers to cover the whole 6--18,GHz frequency range. To achieve a low phase noise performance, a sub-sampling PLL with non-dividers was employed. The synthesizer can achieve phase noise $-$113.7,dBc/Hz@100,kHz in the best case and the reference spur is below $-$60,dBc. The core of the synthesizer consumes about 110,mA*1.2,V.
Kexin QIAO Lei HU Siwei SUN Xiaoshuang MA Haibin KAN
Counting the number of differentially active S-boxes is of great importance in evaluating the security of a block cipher against differential attack. Mouha et al. proposed a technique based on Mixed-Integer Linear Programming (MILP) to automatically calculate a lower bound of the number of differentially active S-boxes for word-oriented block ciphers, and applied it to symmetric ciphers AES and Enocoro-128v2. Later Sun et al. extended the method by introducing bit-level representations for S-boxes and new constraints in the MILP problem, and applied the extended method to PRESENT-80 and LBlock. This kind of methods greatly depends on the constraints in the MILP problem describing the differential propagation of the block cipher. A more accurate description of the differential propagation leads to a tighter bound on the number of differentially active S-boxes. In this paper, we refine the constraints in the MILP problem describing XOR operations, and apply the refined MILP modeling to determine a lower bound of the number of active S-boxes for the Lai-Massey type block cipher FOX in the model of single-key differential attack, and obtain a tighter bound in FOX64 than existing results. Experimental results show that 6, instead of currently known 8, rounds of FOX64 is strong enough to resist against basic single-key differential attack since the differential characteristic probability is upper bounded by 2-64, and thus the maximum differential characteristic probability of 12-round FOX64 is upper bounded by 2-128, where 128 is the key-length of FOX64. We also get the lower bound of the number of differentially active S-boxes for 5-round FOX128, and proved the security of the full-round FOX128 with respect to single-key differential attack.
The development of multichannel audio systems has increased the need for multichannel contents. However, the supply of multichannel audio contents is not sufficient for advanced multichannel systems. Therefore, home entertainment manufacturers need upmixing systems, including systems that utilize monaural time-frequency domain information. Therefore, a monaural ambience extraction algorithm based on nonnegative matrix factorization (NMF) has been developed recently. Ambience signals refer to sound components that do not have obvious spatial images, e.g., wind, rain, and diffuse sound. The developed algorithm provides good upmixing performance; however, the algorithm is a batch process and therefore, it cannot be used by home audio manufacturers. In this paper, we propose an on-line monaural ambience extraction algorithm. The proposed algorithm analyzes the dominant components with an on-line NMF algorithm, and extracts the remaining sound as ambience components. Experiments were performed with artificial mixed signals and real music signals, and the performance of the proposed algorithm was compared with the performance of the conventional batch algorithm as a reference. The experimental results show that the proposed algorithm extracts the ambience components as well as the batch algorithm, despite the on-line constraints.
Shigeyuki KOMURO Shigeru KURIYAMA Takao JINNO
Multimedia contents can be enriched by introducing navigation with image codes readable by camera-mounted mobile devices such as smartphones. Data hiding technologies were utilized for embedding such codes to make their appearances inconspicuous, which can reduce esthetic damage on visual media. This article proposes a method of embedding two-dimensional codes into images based on successive color mixture for a blue-color channel. This technology can make the color of codes mimic those used on a cover image, while preserving their readability for current general purpose image sensors.
Hirofumi SHIMIZU Hiromitsu AWANO Masayuki HIROMOTO Takashi SATO
The modeling of random telegraph noise (RTN) of MOS transistors is becoming increasingly important. In this paper, a novel method is proposed for realizing automated estimation of two important RTN-model parameters: the number of interface-states and corresponding threshold voltage shift. The proposed method utilizes a Gaussian mixture model (GMM) to represent the voltage distributions, and estimates their parameters using the expectation-maximization (EM) algorithm. Using information criteria, the optimal estimation is automatically obtained while avoiding overfitting. In addition, we use a shared variance for all the Gaussian components in the GMM to deal with the noise in RTN signals. The proposed method improved estimation accuracy when the large measurement noise is observed.
Yoshikazu FUJISHIRO Takahiko YAMAMOTO Kohji KOSHIJI
This study proposes a novel method for evaluating the transmission characteristics of a three-phase filter using the “Fortescue-mode S-parameters,” which are S-parameters whose variables are transformed into symmetrical coordinates (i.e., zero-/positive-/negative-phase sequences). The behavior of the filter under three-phase current, including its non-symmetry, can be represented by these S-parameters, without regard to frequency. This paper also describes a methodology for creating modal equivalent circuits that reflect Fortescue-mode S-parameters allowing the effects of circuit components on filter characteristics to be estimated. Thus, this method is useful not only for the measurement and evaluation but also for the analysis and design of a three-phase filter. In addition, the physical interpretation of asymmetrical/symmetrical insertion losses and the conversion method based on Fortescue-mode S-parameters are clarified.
Rui WU Yuuki TSUKUI Ryo MINAMI Kenichi OKADA Akira MATSUZAWA
A 60-GHz power amplifier (PA) with a reliability consideration for a hot-carrier-induced~(HCI) degradation is presented. The supply voltage of the last stage of the PA ($V_{{ m PA}}$) is dynamically controlled by an on-chip digitally-assisted low drop-out voltage regulator (LDO) to alleviate HCI effects. A physical model for estimation of HCI degradation of NMOSFETs is discussed and investigated for dynamic operation. The PA is fabricated in a standard 65-nm CMOS process with a core area of 0.21,mm$^{2}$, which provides a saturation power of 10.1,dBm to 13.2,dBm with a peak power-added efficiency~(PAE) of 8.1% to 15.0% for the supply voltage $V_{{ m PA}}$ which varies from 0.7,V to 1.0,V at 60,GHz, respectively.
Gaussian mixture model (GMM) has recently been applied for image registration given its robustness and efficiency. However, in previous GMM methods, all the feature points are treated identically. By incorporating local class features, this letter proposes a multiple Gaussian mixture models (M-GMM) method for image registration. The proposed method can achieve higher accuracy results with less registration time. Experiments on real image pairs further proved the superiority of the proposed method.