We propose an active learning framework for speech recognition that reduces the amount of data required for acoustic modeling. This framework consists of two steps. We first obtain a phone-error distribution using an acoustic model estimated from transcribed speech data. Then, from a text corpus we select a sentence whose phone-occurrence distribution is close to the phone-error distribution and collect its speech data. We repeat this process to increase the amount of transcribed speech data. We applied this framework to speaker adaptation and acoustic model training. Our evaluation results showed that it significantly reduced the amount of transcribed data while maintaining the same level of accuracy.

A 60 GHz direct conversion transceiver which employs amplitude/phase imbalance cancellation technique is newly proposed. By using the proposed technique, the receive path of the transceiver achieves less than 0.2 dB of amplitude error and less than 3 of phase error at 60 GHz bands over a 10 GHz bandwidth, which relaxes the design accuracy required for baluns used in the transceiver. It also employs a simple and fast calibration algorithm to adjust the locking range of the divide-by-3 injection locked divider in the phase locked loop. Fabricated in 90 nm CMOS technology, the transceiver achieves a low power consumption of 230 mW in transmit mode and 173 mW in receive mode. The output spectrum of 1.76 Gsps π/2-BPSK/QPSK modulation shows the excellent distortion and spurious suppression that meet the IEEE802.11ad draft standard.

Compressing a JPEG image twice will greatly decrease the values of some of its DCT coefficients. This effect can be easily detected by statistics methods. To defend this forensic method, we establish a model to evaluate the security and image quality influenced by the re-compression. Base on the model, an optimized adjustment of the DCT coefficients is achieved by Genetic Algorithm. Results show that the traces of double compression are removed while preserving image quality.

This letter proposes a new model of state machine called Finite Virtual State Machine (FVSM). A memory-based architecture and a procedure for generating FVSM implementations from Finite State Machines (FSMs) are presented. FVSM implementations provide advantages in speed over conventional RAM-based FSM implementations. The results of experiments prove the feasibility of this approach.

A novel method for automatically creating an optimum direction-of-arrival (DOA) estimation algorithm for a given radio environment using a genetic algorithm (GA) is proposed. DOA estimation algorithms are generally described by parameters and operators. The performance of a DOA estimation algorithm is evaluated using root mean square error (RMSE) through computer simulations. A GA searches for the combination of parameters and operators that gives the lowest RMSE. Because a GA can treat only bit strings, Polish notation is used to convert bit strings into a DOA estimation algorithm. A computer simulation showed that the proposed method can create a new angle spectrum function. The created angle spectrum function has higher resolution than the Capon method.

Computational Fluid Dynamics (CFD) is used as a common design tool in the aerospace industry. UPACS, a package for CFD, is convenient for users, since a customized simulator can be built just by selecting desired functions. The problem is its computation speed, which is difficult to enhance by using the clusters due to its complex memory access patterns. As an economical solution, accelerators using FPGAs are hopeful candidate. However, the total scale of UPACS is too large to be implemented on small numbers of FPGAs. For cost efficient implementation, partial reconfiguration which dynamically loads only required functions is proposed in this paper. Here, the MUSCL scheme, which is used frequently in UPACS, is selected as a target. Partial reconfiguration is applied to the flux limiter functions (FLF) in MUSCL. Four FLFs are implemented for Turbulence MUSCL (TMUSCL) and eight FLFs are for Convection MUSCL (CMUSCL). All FLFs are developed independently and separated from the top MUSCL module. At start-up, only required FLFs are selected and deployed in the system without interfering the other modules. This implementation has successfully reduced the resource utilization by 44% to 63%. Total power consumption also reduced by 33%. Configuration speed is improved by 34-times faster as compared to full reconfiguration method. All implemented functions achieved at least 17 times speed-up performance compared with the software implementation.

Bistatic synthetic aperture radar (BSAR) with one fixed station (OF-BSAR) can be used in wide area surveillance, ground moving target indication etc. This paper analyzes the spatial variance of OF-BSAR. Analytical expressions of the spatial invariance region in the data space are given. Using these results, we can determine the spatial invariance region in the data set and the imaging area. After that, we give a data blocking scheme for raw data focusing. Numerical simulation verifies the results of this paper.

The Hilbert-Schmidt independence criterion (HSIC) is a kernel-based statistical independence measure that can be computed very efficiently. However, it requires us to determine the kernel parameters heuristically because no objective model selection method is available. Least-squares mutual information (LSMI) is another statistical independence measure that is based on direct density-ratio estimation. Although LSMI is computationally more expensive than HSIC, LSMI is equipped with cross-validation, and thus the kernel parameter can be determined objectively. In this paper, we show that HSIC can actually be regarded as an approximation to LSMI, which allows us to utilize cross-validation of LSMI for determining kernel parameters in HSIC. Consequently, both computational efficiency and cross-validation can be achieved.

This paper investigates the use of cooperative spectrum sensing (CSS) to detect primary user (PU) signals during spectrum sharing between the PU and the secondary user (SU). In particular, we employ a variant of CSS (which achieves space diversity), called weighted gain combining CSS (W-CSS), which has the potential to achieve increased diversity gain and enhance detection performance. In a typical W-CSS system, the SU needs to obtain the PU signal power information in order to set the proper weight value. However, as it is hard for the SU to ascertain whether the PU is present or absent, this is difficult to obtain. To address this problem, a PU signal power estimation algorithm is introduced. In addition, we also analyze the statistics of the estimator and derive the detection probability of the W-CSS when the PU signal power estimation algorithm is applied. The analysis and related simulation results reveal that the detection probability of the proposed W-CSS under time-variant Rayleigh fading asymptotically approaches the detection probability in an additive white Gaussian noise channel as the number of antennas is increased. This also follows results from our Monte Carlo simulations, showing that multiple antenna elements could suppress the effect of Rayleigh fading. In short, the accuracy of the estimation algorithm is affected by channel variation (especially in the case of fast Rayleigh fading). Hence, to address this problem, we employ multiple antenna elements with a square-law combining energy detector in the W-CSS.

In this paper, we study how to automatically classify mathematical expressions written in MathML (Mathematical Markup Language). It is an essential preprocess to resolve analysis problems originated from multi-meaning mathematical symbols. We first define twelve equation classes based on chapter information of mathematics textbooks and then conduct various experiments. Experimental results show an accuracy of 94.75%, by employing the feature combination of tags, operators, strings, and “identifier & operator” bigram.

This paper proposes a virtual layered successive detector with adaptive transmit signal phase rotation for quadrature amplitude modulation (QAM) that enables high speed communication even in downlinks of wireless communication systems. It is shown that the detection performance is degraded when the eigenvalue of a virtual channel becomes close to the power of the additive white Gaussian noise (AWGN). Therefore, adaptive transmit signal phase rotation is introduced for the detector to improve the transmission performance. For the transmit phase rotation, three techniques to search the rotation angles are proposed, which can reduce the feedback information from the receiver to the transmitter. Among the three proposed techniques, the technique called “iterative variable step step search” is shown to achieve the best performance. Actually, it is confirmed by computer simulation that the variable step search makes the detector attain about 17 dB of a gain at the bit error rate (BER) of 10-5 in 42 multiple-input-multiple-output (MIMO) systems.

This paper proposes an asymptotic method for calculating the received intensity of multi-path millimeter waves transmitted over an undulating surface. This method is a generalization of the asymptotic method that the authors previously derived, based on a quartic phase function approximation, from the physical optics integral expression of the received intensity. The applicability of the previous method is limited to ratios of transmission distance (D) to surface undulation wavelength (λs) of roughly less than 2. This is because the method is based on a quartic phase function approximation. In this paper, this limitation is resolved through this method's generalization, which is achieved by incorporating a technique for calculating the diffraction integral with a higher-degree phase function by using the steepest descent technique with a procedure for systematically identifying the active saddles of the phase function. Numerical examples demonstrate that the proposed asymptotic method can attain calculation accuracy comparable with the physical optics method, even in the case of large D/λs values.

In this letter, a timing synchronization scheme is proposed for cooperative WBAN which can be utilized in wireless medical telemetry systems. The approach efficiently exploits the cyclic structure of a presented preamble in order to effectively separate the superposition of cooperative signals. Then, each timing-offset and each channel parameter are estimated in the separated signal. The proposed scheme effectively recovers the original data based on the timing-offset estimates and the channel estimates. The simulation results reveal the excellent performance of the suggested method in terms of MSE and PER.

In this paper, we introduce a construction of 16-QAM sequences based on known binary sequences using multiple sequences, interleaved sequences and Gray mappings. Five kinds of binary sequences of period N are put into the construction to get five kinds of new 16-QAM sequences of period 4N. These resultant sequences have 5-level autocorrelation {0, ±8, ±8N}, where ±8N happens only once each. The distributions of the periodic autocorrelation are also given. These will provide more choices for many applications.

In this paper, we propose a novel texture analysis method capable of estimating multiple primitives, which are elements repetitively arranged to compose a texture, in multi-structured textures. The approach is based on a texture description model that uses mathematical morphology, called the “Primitive, Grain, and Point Configuration (PGPC)” texture model. The estimation of primitives based on the PGPC texture model involves searching the optimal structuring element for primitives according to a size distribution function and a magnification. The proposed method achieves the following two improvements: (1) the simultaneous estimation of a multiple number of primitives in multi-structured textures with a ranking of primitives on the basis of their significance. and (2) the introduction of flexibility in the process of magnification to obtain a higher degree of fitness of large grains. With a computational combination of different primitives, the method provides an ordered priority to denote the significance of elements. The promising performance of the proposed method is experimentally shown by a comparison with conventional methods.

In this paper, we propose a soft-decoding near-ML MIMO demodulation scheme that achieves near optimal performance with fixed and low complexity. Exploiting the regular structure of bit-to-symbol mapping, the proposed scheme performs hard demodulation to find the first candidate symbol for each stream followed by selection of nearby candidate points such that at least one candidate exists for the computation of likelihood information of bit 0 and 1 without intermediate calculation of the Euclidean distance. This demodulation scheme enables an improvement in performance by guaranteeing the existence of candidates and a significant reduction in the number of distance calculations which is a major complexity burden. The performance is evaluated by computer simulation, and computational complexity is also assessed in terms of the number of complex multiplication.

In a downlink multi-user multiple-input multiple-output (MU-MIMO) system, the vector perturbation (VP) method, which is one of the non-linear precoding methods, is a promising technique with which to maximize the channel capacity. In the VP method, the receiver requires the modulo operation to remove the perturbation signal added at the transmitter. However, owing to noise, the modulo operation may not run correctly which makes the soft demapper produce unreliable log-likelihood ratios (LLRs), resulting in a degradation of system throughput. To enhance the throughput performance, we propose a method that expands the basic perturbation interval and adaptively controls the expansion rate according to the modulation and coding scheme (MCS). The optimum expansion rate is derived by link-level simulation and the system throughput is measured by system-level simulation. The system-level simulation results show that the proposed VP method can obtain higher throughput than the conventional VP method.

This letter proposes a reuse method of unit test cases, which characterize internal behaviors of a called function, for enhancing capability of automatic generation of test cases. Existing test case generation tools have limits in finding solutions to the deep call structure of the source code. In our approach, the complex call structure is simplified by reusing unit test cases of called functions. As unit test cases represent the characteristics of the called function, the internal behaviors of called functions are replaced by the test cases. This approach can be applicable to existing test tools for simplifying the process of generation and enhancing their capabilities.

This paper presents a response time acceleration technique in a high-gain capacitive-feedback frontend amplifier (FA) for high output impedance sensors. Using an auxiliary amplifier as a unity-gain buffer, a sample-and-hold capacitor which is used for band-limiting and sampling the FA output is driven at the beginning of the transient response to make the response faster and then it is re-charged directly by the FA output. A condition and parameters for the response time acceleration using this technique while maintaining the noise level unaffected are discussed. Theoretical analysis and simulation results show that the response time can be less than half of the case without the acceleration technique for the specified settling error of less than 0.5%.

In most software development environments, time, computing and human resources needed to perform the testing of a component is strictly limited. In order to deal with such situations, this paper proposes a method of creating the best possible test suite (covering the maximum number of 3-tuples) within a fixed number of test cases.