We propose a new algorithm to suppress both stationary background noise and nonstationary directional interference noise in a speech enhancement system that employs the generalized sidelobe canceller. Our approach builds on advances in generalized sidelobe canceller design involving the transfer function ratio. Our system is composed of three stages. The first stage estimates the transfer function ratio on the acoustic path, from the nonstationary directional interference noise source to the microphones, and the powers of the stationary background noise components. Secondly, the estimated powers of the stationary background noise components are used to execute spectral subtraction with respect to input signals. Finally, the estimated transfer function ratio is used for speech enhancement on the primary channel, and an adaptive filter reduces the residual correlated noise components of the signal. These algorithmic improvements give consistently better performance than the transfer function generalized sidelobe canceller when input signal-to-noise ratio is 10 dB or lower.

This paper investigates the fixed-slope lossy coding of individual sequences and nonstationary sources. We clarify that, for a given individual sequence, the optimal cost attainable by the blockwise lossy encoders is equal to the optimal average cost with respect to the empirical distribution of the given sequence. Moreover, we show that, for a given nonstationary source, the optimal cost attainable by the blockwise encoders is equal to the supremum of the optimal average cost over all the stationary sources in the stationary hull of the given source. In addition, we show that the universal lossy coding algorithm based on Lempel-Ziv 78 code attains the optimal cost for any individual sequence and any nonstationary source.

In this Letter, we propose a new adaptive step-size widely linear constant modulus algorithm (CMA) in DS-CDMA systems especially for time-varying interference environments. The widely linear estimation enables CMA to produce better output signal to interference plus noise ratio (SINR) and the adaptive step-size tackles the time-varying interference environment effectively. The simulations confirm that the proposed algorithm shows better performance in a DS-CDMA system employing a BPSK modulation than other algorithms without use of widely linear processing.

A block-matching-based self-organizing map (BMSOM) is presented. Finding a winner is carried out for each block, which is a set of neurons arranged in square. The proposed learning process updates the reference vectors of all of the neurons in a winner block. Then, the degrees of vector modifications are mainly controlled by the size (i.e., the number of neurons) of the winner block. To prevent a single cluster with neurons from splitting into some disjointed clusters, the restriction of the block size is imposed in the beginning of learning. At the main stage, this restriction is canceled. In BMSOM learning, the size of a winner block does not always decrease monotonically. The formula used to update the reference vectors is basically uncontrolled by time. Therefore, even if a map is in a nonstationary environment, training the map is probably pursued without interruption to adjust time-controlled parameters such as learning rate. Numerical results demonstrate that the BMSOM makes it possible to improve the plasticity of maps in a nonstationary environment and incremental learning.

A wide variety of visual textures could be successfully modeled as spatially variant by quantitatively describing them through the variation of their local spatial frequency and/or local orientation components. This class of patterns includes flow-like, granular or oriented textures. Modeling is achieved by assuming that locally, textured images contain a single dominant component describing their local spatial frequency and modulating amplitude or contrast. Spatially variant textures are non-homogeneous in the sense of having nonstationary local spectra, while remaining locally coherent. Segmenting spatially variant textures is the challenging task undertaken in this paper. Usually, the goal of texture segmentation is to split an image into regions with homogeneous textural properties. However, in the case of image regions with spatially variant textures, there is no global homogeneity present and thus segmentation passes through identification of regions with globally nonstationary, but locally coherent, textural content. Local spatial frequency components are accurately estimated using Gabor wavelet outputs along with the absolute magnitude of the convolution of the input image with the first derivatives of the underlying Gabor function. In this paper, a frequency estimation approach is used for segmentation. Indeed, at the boundary between adjacent textures, discontinuities occur in texture local spatial frequency components. These discontinuities are interpreted as corresponding to texture boundaries. Experimental results are in remarkable agreement with human visual perception, and demonstrate the effectiveness of the proposed technique.

This paper is, in half part, written in review nature, and presents recent theoretical results on linear-filtering and -prediction problems of nonstationary Gaussian processes. First, the basic concepts, signal and noise, are mathematically characterized, and information sources are defined by linear stochastic differential equations. Then, it is shown that the solution to a conventional problem of filtering or prediction of a nonstationary time series is, in principle, reducible to a problem, of which solution is given by Kalman-Bucy's theory, if one can solve a problem of finding the canonical representation of a Gaussian process such that it has the same covariance functions as those of the time series under consideration. However, the problem mentioned above is left open. Further, the problem of time-frequency analysis is discussed, and physical realizability of the evolutionary, i.e., the online, spectral analyzer is shown. Methods for dealing with differential operators are presented and their basic properties are clarified. Finally, some of related open problems are proposed.

We propose a learning algorithm for self-organizing neural networks to form a topology preserving map from an input manifold whose topology may dynamically change. Experimental results show that the network using the proposed algorithm can rapidly adjust itself to represent the topology of nonstationary input distributions.

We present a mechanism, named the law of the jungle (LOJ), to improve the Kohonen learning. The LOJ is used to be an adaptive vector quantizer for approximating nonstationary probability distribution functions. In the LOJ mechanism, the probability that each node wins in a competition is dynamically estimated during the learning. By using the estimated win probability, "strong" nodes are increased through creating new nodes near the nodes, and "weak" nodes are decreased through deleting themselves. A pair of creation and deletion is treated as an atomic operation. Therefore, the nodes which cannot win the competition are transferred directly from the region where inputs almost never occur to the region where inputs often occur. This direct "jump" of weak nodes provides rapid convergence. Moreover, the LOJ requires neither time-decaying parameters nor a special periodic adaptation. From the above reasons, the LOJ is suitable for quick approximation of nonstationary probability distribution functions. In comparison with some other Kohonen learning networks through experiments, only the LOJ can follow nonstationary probability distributions except for under high-noise environments.

An analysis method is proposed for nonstationary waveforms. Modelling of a nonstationary waveform is first given in this paper. A waveform is represented by multiple oscillations. The instantaneous phase angle of each oscillation is written by three terms, predictive component, residual component, and initial phase constant. By this modelling, waveform analysis results in estimations of frequency, calculation of residual pbase in instantaneous phase angle. The Instantaneous Maximum Entropy Methods (IMEN) is utilized for frequency estimation. The residual phase angle is obtained by the Vandermonde matrix and the condition of continuity of phase angle among n-neighbourhood. Another analysis method is also proposed by the normalization of waveform parameters. The evaluation of the proposed method is done using artificially composed waveform signals. Novel and useful knowledge was provided by this analysis.

In recent years, fractal processes have played important roles in various application fields. Since a 1/f process possesses the statistical self-similarity, it is considered sa a main part of fractal signal modeling. On the other hand, noise reduction is often needed in real-world signal processing. Hence, we propose an enhancement algorithm for 1/f signal disturbed by white noise. The algorithm is based on constrained minimization in a wavelet domain: the power of 1/f signal distortion in the wavelet domain is minimized under a constraint that the power of residual noise in the wavelet domain is smaller than a threshold level. We solve this constrained minimization problem using a Lagrangian equation. We also consider a setting method of the Lagrange multiplier in the proposed algorithm. In addition, we will confirm that the proposed algorithm with this Lagrange multiplier setting method obtains better enhancement results than the conventional algorithm through computer simulations.

In the correspondence discrete Wigner higher order spectra (WHOS) of harmonizable random signals are addressed and their relations with polyspectra (HOS) are illustrated. It is shown, that discrete WHOS of a random stationary signal do not reduce to the aliased polyspectra in a similar way as Wigner distribution (WD) reduces to the power spectrum of a random signal. Wigner 2nd-order time-frequency distribution of deterministic signals and the 3rd-order spectrum of stationary signals are presented in their modified forms to be used to estimate time-varying third-order spectrum of discrete nonstationary random harmonizable processes.

A new trial of statistical evaluation for an output response of power linear type acoustic systems with nonstationary random input is proposed. The purpose of this study is to predict the output probability distribution function on the basis of a standard type pre-experiment in a laboratoty. The statistical properties like nonstationarity, non-Gamma distribution property and various type linear and non-linear correlations of input signal are reflected in the form of differential operation with respect to distribution parameters. More concretely, the pre-experiment is carried out for a power linear acoustic system excited only by the Gamma distribution type sandard random input. Considering the non-negative random property for the output response of a power linear system, the well-known statistical Laguerre expansion series type probability expression is first employed as the framework of basic probability distribution expression on the output power fluctuation. Then, the objective output probability distribution for a non-stationary case can be easily derived only by successively employing newly introduced differential operators to this basic probability distribution of statistical Laguerre expansion series type. As an application to the actual noise environment, the proposed method is employed for an evaluation problem on the stochastic response probability distribution for an acoustic sound insulation system excited by a nonstationary input noise.

A ner trial of statistical evaluation for a nonstationary traffic flow and its traffic noise is proposed as a prediction method of its probability distribution function by considering the temporal change of distribution parameters especially from a structural viewpoint. First, a headway distribution of the nonstationary traffic flow passing through within a road segment is proposed on the basis of an Erlang distribution by reflecting a temporal change of its distribution parameters. Then, an initial phase density concerning with asynchronous counting method and the probability of counting n cars over a long time interval are derived from the above nonstationary expression of headway distribution. Thus, the statistics of noise intensity at an observation point has been predicted by combining the above probabilistic factors and deterministic factors related to noise propagation environment with use of a compound stochastic process model. Finally, te effectivenss of the proposed theory has been confirmed experimentally by applying it to the actual traffic flow on a highway.