Showing papers in "IEEE Signal Processing Letters in 1994"
TL;DR: An efficient method is proposed to obtain a good initial codebook that can accelerate the convergence of the generalized Lloyd algorithm and achieve a better local minimum as well.
Abstract: The generalized Lloyd algorithm plays an important role in the design of vector quantizers (VQ) and in feature clustering for pattern recognition. In the VQ context, this algorithm provides a procedure to iteratively improve a codebook and results in a local minimum that minimizes the average distortion function. We propose an efficient method to obtain a good initial codebook that can accelerate the convergence of the generalized Lloyd algorithm and achieve a better local minimum as well. >
374 citations
TL;DR: This work proposes a "probing" mode during which probing signals received at the mobiles are fed back to the transmitter, enabling the transmitter to form the necessary transmission beampatterns.
Abstract: We address the problem of transmitting multiple cochannel signals from an antenna array to several receivers so that each receiver gets its intended signal with minimum crosstalk from the remaining signals. In addition to the usual "information" mode, we propose a "probing" mode during which probing signals received at the mobiles are fed back to the transmitter. These probing signals are used to identify an unknown propagation environment, enabling the transmitter to form the necessary transmission beampatterns. >
258 citations
TL;DR: The authors exploit the temporal structure of the digital signals to simultaneously determine the array response and the bit sequence for each signal to propose a novel approach for separating and estimating multiple co-channel digital signals using an antenna array.
Abstract: Proposes a novel approach for separating and estimating multiple co-channel digital signals using an antenna array. The spatial response of the array is unknown. The authors exploit the temporal structure of the digital signals to simultaneously determine the array response and the bit sequence for each signal. Uniqueness of the estimates is established for signals with BPSK modulation format. This new approach is applicable to an unknown array geometry and propagation environment, which is particularly useful in digital mobile communications. Simulation results demonstrate its promising performance. >
199 citations
TL;DR: A derivation of the normalized LMS algorithm is generalized, resulting in a family of projection-like algorithms based on an L/sub p/-minimized filter coefficient change, which include the simplified NLMS algorithm of Nagumo and Noda (1967) and an even simpler single-coefficient update algorithm based on the maximum absolute value datum of the input data vector.
Abstract: A derivation of the normalized LMS algorithm is generalized, resulting in a family of projection-like algorithms based on an L/sub p/-minimized filter coefficient change. The resulting algorithms include the simplified NLMS algorithm of Nagumo and Noda (1967) and an even simpler single-coefficient update algorithm based on the maximum absolute value datum of the input data vector. A complete derivation of the algorithm family is given, and simulations are performed to show the convergence behaviors of the algorithms. >
190 citations
TL;DR: A large class of physical phenomena observed in practice exhibit non-Gaussian behavior, and the /spl alpha/-stable distributions, which have heavier tails than Gaussian distributions, are considered to model non- Gaussian signals.
Abstract: A large class of physical phenomena observed in practice exhibit non-Gaussian behavior. In the letter /spl alpha/-stable distributions, which have heavier tails than Gaussian distributions, are considered to model non-Gaussian signals. Adaptive signal processing in the presence of such a noise is a requirement of many practical problems. Since direct application of commonly used adaptation techniques fail in these applications, new algorithms for adaptive filtering for /spl alpha/-stable random processes are introduced. >
106 citations
TL;DR: A program is presented that is superior in speed and accuracy to the best methods to the authors' knowledge, i.e., Jenkins/Traub (1975) program and the eigenvalue method to improve the accuracy for spectral factorization in the case that there are double roots on the unit circle.
Abstract: Finding polynomial roots rapidly and accurately is an important problem in many areas of signal processing. We present a program that is superior in speed and accuracy to the best methods to our knowledge, i.e., Jenkins/Traub (1975) program and the eigenvalue method. Based on this, we give a simple approach to improve the accuracy for spectral factorization in the case that there are double roots on the unit circle. >
92 citations
TL;DR: A blind symbol estimation technique for digital communication is developed by exploiting a special data structure of the oversampled channel output to achieve direct symbol estimation without determining the channel characteristics.
Abstract: A blind symbol estimation technique for digital communication is developed by exploiting a special data structure of the oversampled channel output. The proposed method achieves direct symbol estimation without determining the channel characteristics. Moreover, if the transmitting symbols belong to a finite set of alphabets, the new approach can be extended to handle multiple sources. >
89 citations
TL;DR: A dual form of the Wigner higher order spectra is introduced and an efficient distribution for time-frequency signal analysis (L-Wigner distribution) is derived.
Abstract: A dual form of the Wigner higher order spectra is introduced. Its analysis in the case of multicomponent signals is performed. An efficient distribution for time-frequency signal analysis (L-Wigner distribution) is derived from that analysis. The theory is illustrated on a numerical example. >
84 citations
TL;DR: The author proves that the matrixing operations in MPEG subband filtering can be efficiently computed using fast 32-point DCT or inverse DCT (IDCT) algorithms.
Abstract: Subband filtering is one of the most compute-intensive operations in the MPEG audio coding standard. The author proves that the matrixing operations in MPEG subband filtering can be efficiently computed using fast 32-point DCT or inverse DCT (IDCT) algorithms. >
79 citations
TL;DR: In this paper, the impact of nonGaussian impulsive noise combined with Gaussian noise on the performance of binary transmission is analyzed, where the impulsive noises are modeled as an alpha-stable process and the probability of error for optimum, linear and nonlinear receivers is derived.
Abstract: The impact of nonGaussian impulsive noise combined with Gaussian noise on the performance of the binary transmission is analyzed The impulsive noise is modeled as an alpha-stable process The probability of error for optimum, linear and nonlinear receivers is derived The proposed nonlinear detectors show substantial improvements in performance compared to linear ones The obtained results will be useful in performance evaluation of digital communication links subject to Gaussian and impulsive noises >
77 citations
TL;DR: Theoretical and experimental results regarding both approximation error and speed improvement prove the validity of the proposed algorithm.
Abstract: The vector median filter has good filtering capabilities; nevertheless, its huge computational complexity significantly limits its practical usability. A vector median filter based on a fast approximation of the Euclidean norm is presented. The proposed algorithm couples computational and filtering effectiveness, and it is well suited for hardware implementation. Theoretical and experimental results regarding both approximation error and speed improvement prove the validity of the proposed algorithm. >
TL;DR: Fast algorithms for the evaluation of running windowed Fourier and continuous wavelet transforms are presented and approximate complex-modulated Gaussians as closely as desired and may be optimally localized in time and frequency.
Abstract: Fast algorithms for the evaluation of running windowed Fourier and continuous wavelet transforms are presented. The analysis functions approximate complex-modulated Gaussians as closely as desired and may be optimally localized in time and frequency. The Gabor filtering is performed indirectly by convolving a premodulated signal with a Gaussian-like window and demodulating the output. The window functions are either B-splines dilated by an integer factor m or quasi-Gaussians of arbitrary size generated from the n-fold convolution of a symmetrical exponential. Both approaches result in a recursive implementation with a complexity independent of the window size (O(N)). >
TL;DR: The authors show that making use of the discrete wavelet transform to analyse data implies performing a preliminary initialization of the fast pyramidal algorithm, and an approximation enabling easy performance is proposed.
Abstract: The authors show that making use of the discrete wavelet transform to analyse data implies performing a preliminary initialization of the fast pyramidal algorithm. An approximation enabling easy performance of such an initialization is proposed. >
TL;DR: A blind channel identification scheme based on oversampling the channel output and employing a (discrete) cyclic correlation function similar to Gardner (1991) is proposed and it is shown that the known results on blind identifiability follow directly.
Abstract: A blind channel identification scheme based on oversampling the channel output and employing a (discrete) cyclic correlation function similar to Gardner (1991) is proposed. It is shown that the known results on blind identifiability follow directly. Further, this connection suggests a new and efficient time-domain algorithm for blind channel identification that is applied to some representative examples. >
TL;DR: It is demonstrated that two previously proposed methods for combining the information content from multiple spectrograms into a single, positive time-frequency function are optimal in a cross-entropy sense.
Abstract: We demonstrate that two previously proposed methods for combining the information content from multiple spectrograms into a single, positive time-frequency function are optimal in a cross-entropy sense. The goal in combining the spectrograms is to obtain an improved approximation of the joint time-frequency signal density by overcoming limitations of any single spectrogram. An example of each method is provided, and results are compared with spectrograms and a Cohen-Posch (1985) time-frequency density (TFD) of a nonstationary pulsed tone signal. The proposed combinations are effective and can be efficiently computed. >
TL;DR: The minimax near-field design problem of a broadband beamformer is solved as a quadratic programming formulation of the weighted Chebyshev approximation problem and the method can be applied to the design of multidimensional digital FIR filters with an arbitrarily specified amplitude and phase.
Abstract: A method to solve a general broadband beamformer design problem is formulated as a quadratic program. As a special case, the minimax near-field design problem of a broadband beamformer is solved as a quadratic programming formulation of the weighted Chebyshev approximation problem. The method can also be applied to the design of multidimensional digital FIR filters with an arbitrarily specified amplitude and phase. For linear phase multidimensional digital FIR filters, the quadratic program becomes a linear program. Examples are given that demonstrate the minimax near-field behavior of the beamformers designed. >
TL;DR: The decomposition of the characteristic waveform is decomposed into a slowly evolving waveform and a rapidly evolving waveforms, representing the quasi-periodic and other components of speech, respectively, which allows efficient coding of voiced and unvoiced speech at bit rates between 2 and 8 kb/s.
Abstract: The speech signal is represented by an evolving characteristic waveform. The characteristic waveform is decomposed into a slowly evolving waveform and a rapidly evolving waveform, representing the quasi-periodic and other components of speech, respectively. These two evolving waveforms have fundamentally different quantization requirements. The decomposition allows efficient coding of voiced and unvoiced speech at bit rates between 2 and 8 kb/s. >
TL;DR: The authors first demonstrate that the forward and inverse discrete cosine transform (DCT, IDCT) can be represented by Chebyshev polynomials of the third and second kind, respectively, and derive recursive algorithms for the DCT and IDCT with arbitrary length from the recursive formulae for the Chebyshemials.
Abstract: The authors first demonstrate that the forward and inverse discrete cosine transform (DCT, IDCT) can be represented by Chebyshev polynomials of the third and second kind, respectively. Then, they derive recursive algorithms for the DCT and IDCT with arbitrary length from the recursive formulae for the Chebyshev polynomials. The proposed algorithms are particularly suitable for VLSI implementation using array processing architectures. >
TL;DR: The authors propose the use of a radial basis function (RBF) network for direction-of-arrival (DOA) estimation and results show that the new technique has a better performance in terms of estimation errors than the standard MUSIC algorithm.
Abstract: The authors propose the use of a radial basis function (RBF) network for direction-of-arrival (DOA) estimation. The RBF network is used to approximate the functional relationship between sensor outputs and the direction of arrivals. Simulation results show that the new technique has a better performance in terms of estimation errors than the standard MUSIC algorithm. >
TL;DR: A new method for VFR using the norm of the derivative parameters in deciding to retain or to discard a frame is introduced, and informal inspection of speech spectrograms shows that this new method puts more emphasis on the transient regions of the speech signal.
Abstract: Variable frame rate (VFR) analysis is a technique used in speech processing and recognition for discarding frames that are too much alike. The article introduces a new method for VFR. Instead of calculating the distance between frames, the norm of the derivative parameters is used in deciding to retain or to discard a frame, informal inspection of speech spectrograms shows that this new method puts more emphasis on the transient regions of the speech signal. Experimental results with a hidden Markov model (HMM) based system show that the new method outperforms the classical method. >
TL;DR: Using threshold decomposition, it is shown that median filtering operation minimizes a two-term cost function of the output state of the median filter.
Abstract: In this letter, we use a new approach for studying the properties of median filtering. Specifically, using threshold decomposition, it is shown that median filtering operation minimizes a two-term cost function of the output state of the median filter. The first term of the cost function measures the smoothness between the median filter output and its neighbor points within the operation window, and the second term measures the discrepancy between the filter output and its original signal. The results from this study have provided us with a new tool to analyze and understand some of the properties of the median filtering operation, including weighted median filtering. >
TL;DR: The approach to decrease the acoustic mismatch between a test utterance Y and a given set of speech hidden Markov models /spl Lambda//sub X/ to reduce the recognition performance degradation caused by possible distortions in the test utterances is presented.
Abstract: Presents an approach to decrease the acoustic mismatch between a test utterance Y and a given set of speech hidden Markov models /spl Lambda//sub X/ to reduce the recognition performance degradation caused by possible distortions in the test utterance. This is accomplished by a parametric function that transforms either U or /spl Lambda//sub X/ to better match each other. The functional form of the transformation depends on prior knowledge about the mismatch, and the parameters are estimated along with the recognized string in a maximum-likelihood manner. experimental results verify the efficacy of the approach in improving the performance of a continuous speech recognition system in the presence of mismatch due to different transducers and transmission channels. >
TL;DR: An improved implementation of the oddly stacked Princen-Bradley filter bank is proposed, having easy and elegant merging of output rotations and windows and reduced arithmetic complexity.
Abstract: An improved implementation of the oddly stacked Princen-Bradley filter bank is proposed. The new implementation is based on the same approach as suggested by Duhamel et al. [1991], where the complex fast Fourier transform algorithm of length equal to half of the number of subbands is used in order to obtain a significant reduction of number of arithmetic operations. In the implementation, the analysis filter bank is similar to Duhamel's, with some minor differences. However, the synthesis filter bank is quite different, having easy and elegant merging of output rotations and windows and reduced arithmetic complexity. >
TL;DR: Estimation of various Markov-modulated time series including the Markov chain transition probabilities and the time-series coefficients using the expectation maximization (EM) algorithm and the recursive EM algorithm to obtain on-line parameter estimates.
Abstract: We consider the estimation of various Markov-modulated time series. We obtain maximum likelihood estimates of the time-series parameters including the Markov chain transition probabilities and the time-series coefficients using the expectation maximization (EM) algorithm. In addition, the recursive EM algorithm is used to obtain on-line parameter estimates. Simulation studies show that both algorithms yield satisfactory results. >
TL;DR: A performance analysis of the dual sign algorithm is considered when the signals are zero-mean stationary Gaussian, and the additive noise of the desired response is zero- mean stationary contaminated-Gaussian (CG).
Abstract: In the previous analysis of the dual sign algorithm (DSA), Gaussian signals were assumed. When the desired response contains additive impulsive interference, however, the analysis seems to be inadequate. In this article, a performance analysis of the DSA is considered when the signals are zero-mean stationary Gaussian, and the additive noise of the desired response is zero-mean stationary contaminated-Gaussian (CG). Through computer simulations, our analysis is validated. It is also shown that the DSA is less vulnerable to impulsive interference than the least-mean square (LMS) algorithm. >
TL;DR: It is shown how Tong's method can be simply modified so that the unknown correlation coefficient can be estimated and then compensated, which leads to an improved performance.
Abstract: Presents a source correlation compensation approach for the blind channel identification method proposed by Tong et al. [1991, 1994]. Instead of assuming independent and identically distributed (i.i.d.) source symbols as assumed by Tong et al., the present authors consider the case where each source symbol is weakly correlated to its nearest neighbors with an unknown correlation coefficient. They show how Tong's method can be simply modified so that the unknown correlation coefficient can be estimated and then compensated, which leads to an improved performance. >
TL;DR: This letter presents a simple technique that uses both spatial and temporal information to improve signal copy performance and uses an initial blind estimate of the signals to compute a least-squares estimates of the array response, which in turn is used to update the signal estimates.
Abstract: Conventional methods for signal copy require one to estimate the directions of arrival (DOA's) of the signals prior to computing the weight vectors. Blind copy algorithms alleviate the need for DOA estimation (and hence the need for array calibration data) by exploiting the temporal rather than spatial structure of the signals, but they converge slowly in some cases. In this letter, we present a simple technique that uses both spatial and temporal information to improve signal copy performance. Specifically, the algorithm uses an initial blind estimate of the signals to compute a least-squares estimate of the array response, which in turn is used to update the signal estimates. >
TL;DR: Simulation results show that the proposed PVQ design outperforms the conventional PVQ schemes, such as the closed-loop design and the jointly-optimized technique.
Abstract: A joint optimization technique is developed for designing the predictor and quantizer of a predictive vector quantizer (PVQ). The proposed technique is based on a constrained optimization technique that makes use of a Lagrangian formulation and iteratively solves the Lagrangian error function to obtain a locally optimal solution for the predictor and quantizer. Simulation results show that the proposed PVQ design outperforms the conventional PVQ schemes, such as the closed-loop design and the jointly-optimized technique. >
TL;DR: Speaker identification improved from 55 to 70% correct classification when the full set of new resonant energy-based features were added as an independent stream to conventional mel-cepstra.
Abstract: Onset times of resonant energy pulses are measured with the high-resolution Teager operator and used as features in the Reynolds Gaussian-mixture speaker identification algorithm. Feature sets are constructed with primary pitch and secondary pulse locations derived from low and high speech formants. Preliminary testing was performed with a confusable 40-speaker subset from the NTIMIT (telephone channel) database. Speaker identification improved from 55 to 70% correct classification when the full set of new resonant energy-based features were added as an independent stream to conventional mel-cepstra. >
TL;DR: A simple method is introduced, which they call dynamic M-search, that is both fast and effective for searching the residual vector quantization tree and is applied to image coding.
Abstract: The encoding procedure for multistage residual vector quantizers consists of searching a tree to find the codeword that best matches the input. Many techniques have been proposed for searching such trees, ranging from the slow and faithful to the fast and inaccurate. The present authors introduce a simple method, which they call dynamic M-search, that is both fast and effective for searching the residual vector quantization tree. The new method is applied to image coding, performance results are presented, and advantages are shown over conventional M-search. >