Showing papers in "IEEE Signal Processing Letters in 1997"
TL;DR: This article shows that the infomax (information-maximization) principle is equivalent to the maximum likelihood.
Abstract: Algorithms for the blind separation of sources can be derived from several different principles. This article shows that the infomax (information-maximization) principle is equivalent to the maximum likelihood. The application of the infomax principle to source separation consists of maximizing an output entropy.
739 citations
TL;DR: A mew generalized block-edge impairment metric (GBIM) is presented in this paper as a quantitative distortion measure for blocking artifacts in digital video and image coding and is found to be consistent with subjective evaluation.
Abstract: A mew generalized block-edge impairment metric (GBIM) is presented in this paper as a quantitative distortion measure for blocking artifacts in digital video and image coding. This distortion measure does not require the original image sequence as a comparative reference, and is found to be consistent with subjective evaluation.
376 citations
TL;DR: A progressive image compression scheme whose performance on a noisy channel is significantly better than that of previously known techniques and effectively no degradation due to channel noise can be detected.
Abstract: We cascade an existing image coder with carefully chosen error control coding, and thus produce a progressive image compression scheme whose performance on a noisy channel is significantly better than that of previously known techniques. The main idea is to trade off the available transmission rate between source coding and channel coding in an efficient manner. This coding system is easy to implement and has acceptably low complexity. Furthermore, effectively no degradation due to channel noise can be detected; instead, the penalty paid due to channel noise is a reduction in source coding resolution. Detailed numerical comparisons are given that can serve as benchmarks for comparisons with future encoding schemes. For example, for the 512/spl times/512 Lena image, at a transmission rate of 1 b/pixel, and for binary symmetric channels with bit error probabilities 10/sup -3/, 10/sup -2/, and 10/sup -1/, the proposed system outperforms previously reported results by at least 2.6, 2.8, and 8.9 dB, respectively.
342 citations
TL;DR: Multirate precoding using filterbanks induces cyclo-stationarity at the transmitter and guarantees blind identifiability of frequency selective communication channels with minimal decrease of information rate and without restrictions on zero locations.
Abstract: Multirate precoding using filterbanks induces cyclo-stationarity at the transmitter and guarantees blind identifiability of frequency selective communication channels with minimal decrease of information rate and without restrictions on zero locations. Finite impulse response (FIR) filterbank decoders are capable of equalizing blindly (and perfectly in the absence of noise) FIR channels without constraints on their zeros.
212 citations
TL;DR: The consider the problem of recovering p synchronous communication signals that are transmitted through a multiple-input/multiple-output (MIMO) linear channel and are, therefore, received in the presence of both interuser (IUI) and intersymbol interference (ISI).
Abstract: The consider the problem of recovering p synchronous communication signals that are transmitted through a multiple-input/multiple-output (MIMO) linear channel and are, therefore, received in the presence of both interuser (IUI) and intersymbol interference (ISI). A multichannel linear equalization approach is taken, and we propose to adjust the equalizer coefficients with a blind adaptive algorithm (without the use of training data). This multiuser constant modulus algorithm (MU-CMA) is derived from the minimization of a cost function that penalizes deviations of the equalized signals from the constant modulus property as well as cross-correlations between them. The proposed scheme appears to be an appealing technique for multiuser blind equalization that combines good convergence properties with low computational complexity.
211 citations
TL;DR: A new member of the family of mixed-norm stochastic gradient adaptive filter algorithms for system identification applications based upon a convex function of the error norms that underlie the least mean square (LMS) and least absolute difference (LAD) algorithms is proposed.
Abstract: We propose a new member of the family of mixed-norm stochastic gradient adaptive filter algorithms for system identification applications based upon a convex function of the error norms that underlie the least mean square (LMS) and least absolute difference (LAD) algorithms. A scalar parameter controls the mixture and relates, approximately, to the probability that the instantaneous desired response of the adaptive filter does not contain significant impulsive noise. The parameter is calculated with the complementary error function and a robust estimate of the standard deviation of the desired response. The performance of the proposed algorithm is demonstrated in a system identification simulation with impulsive and Gaussian measurement noise.
201 citations
TL;DR: The purpose of this paper is to introduce extensions of the FT¿s convolution theorem, dealing with the FRFT of a product and of a convolution of two functions.
Abstract: The fractional Fourier transform (FRFT) is a generalization of the classical Fourier transform (FT). It has recently found applications in several areas, including signal processing and optics. Many properties of this transform are already known, but an extension of the FT?s convolution theorem is still missing. The purpose of this paper is to introduce extensions of this theorem, dealing with the FRFT of a product and of a convolution of two functions.
150 citations
TL;DR: A detector of a spatially distributed target in white Gaussian noise using a simple detector form, whose detection performance is robust over different scattering densities.
Abstract: A detector of a spatially distributed target in white Gaussian noise is developed. A reasonable distribution for the a priori target scatterer density is assumed, and a detector that incorporates this a priori knowledge is given. A simple detector form results, whose detection performance is robust over different scattering densities.
126 citations
TL;DR: It is shown that for a generic two-component AM-FM signal, the interpretation of instantaneous frequency holds only when the components are of equal strength.
Abstract: Instantaneous frequency, taken as the derivative of the phase of the signal, is interpreted in the time-frequency literature as the average frequency of the signal at each time. We point out some difficulties with this interpretation, and show that for a generic two-component AM-FM signal, the interpretation holds only when the components are of equal strength. We conclude that instantaneous frequency and the average frequency at each time are generally two different quantities. One possible interpretation of the difference between these two quantities is suggested.
105 citations
TL;DR: A critical wavelet coefficient quantification is defined, i.e., the coarsest quantification that allows perfect reconstruction for the Haar transform and for arbitrarily smooth wavelet transforms derived from it, allowing implementation of multiresolution subband compression schemes.
Abstract: Signal compression can be obtained by wavelet transformation of integer input data followed by quantification and coding. As the quantification is usually lossy, the whole compression/decompression scheme is lossy too. We define a critical wavelet coefficient quantification, i.e., the coarsest quantification that allows perfect reconstruction. This is demonstrated for the Haar transform and for arbitrarily smooth wavelet transforms derived from it. The new integer wavelet transform allows implementation of multiresolution subband compression schemes, in which the decompressed data are gradually refined, retaining the option of perfect reconstruction.
98 citations
TL;DR: In this paper, the authors proposed a new algorithm for both vector quantizer design and clustering analysis as an alternative to the conventional K-means algorithm, which converges to a better locally optimal codebook with an accelerated convergence speed.
Abstract: The K-means algorithm is widely used in vector quantizer (VQ) design and clustering analysis In VQ context, this algorithm iteratively updates an initial codebook and converges to a locally optimal codebook in certain conditions It iteratively satisfies each of the two necessary conditions for an optimal quantizer; the nearest neighbor condition for the partition and centroid condition for the codevectors In this letter, we propose a new algorithm for both vector quantizer design and clustering analysis as an alternative to the conventional K-means algorithm The algorithm is almost the same as the K-means algorithm except for a modification at codebook updating step It does not satisfy the centroid condition iteratively, but asymptotically satisfies it as the number of iterations increases Experimental results show that the algorithm converges to a better locally optimal codebook with an accelerated convergence speed
TL;DR: A new least-squares-based approach for the joint diagonalization problem arising in blind beamforming is presented and the resulting estimation criterion turns out to coincide with that proposed by Cardoso and Souloumaic on intuitive grounds, thus establishing the optimality of their criterion in the least-Squares (LS) sense.
Abstract: We present a new least-squares-based approach for the joint diagonalization problem arising in blind beamforming. The resulting estimation criterion turns out to coincide with that proposed by Cardoso and Souloumaic (see IEE Proc. F, Radar Signal Process., vol.140, no.6, p.362-70, Dec. 1993) on intuitive grounds, thus establishing the optimality of their criterion in the least-squares (LS) sense.
TL;DR: The two aforementioned criteria are proved to be included in the wide class of contrast functions, which is here defined through simple conditions, and many other contrast functions may be considered.
Abstract: Two contrasts for the problem of multichannel blind deconvolution have been given and theoretically studied by Comon [1996]. The maximization of these criteria allows us to solve the problem of multi-input/multi-output (MIMO) blind deconvolution. In this paper, we show that many other contrast functions may be considered. The two aforementioned criteria are proved to be included in the wide class of contrast functions, which is here defined through simple conditions.
Journal Article•
TL;DR: These vector fields are shown to be connected with both an intrinsic phase characterization and a scalar potential, which allows for the generalization of the original reassignment process to a differential version based on a dynamical evolution of time-frequency particles.
Abstract: A geometrical description is given for reassignment vector fields of spectrograms. These vector fields are shown to be connected with both an intrinsic phase characterization and a scalar potential. This allows for the generalization of the original reassignment process to a differential version based on a dynamical evolution of time-frequency particles.
TL;DR: This work designs tenth-order IIR filters, suitable for efficient real-time implementation, from preprocessed HRTF impulse responses that are of significantly superior quality to current IIR models derived with the Prony and Yule-Walker methods.
Abstract: We propose a novel technique for the design of low-order infinite impulse response (IIR) filter models of head-related transfer functions (HRTFs) that uses balanced model truncation. We design tenth-order IIR filters, suitable for efficient real-time implementation, from preprocessed HRTF impulse responses that are of significantly superior quality to current IIR models derived with the Prony and Yule-Walker methods.
TL;DR: A fast encoding algorithm for vector quantization that uses two characteristics of a vector, mean, and variance simultaneously to save computation time all the more.
Abstract: In this letter, we present a fast encoding algorithm for vector quantization that uses two characteristics of a vector, mean, and variance. Although a similar method using these features was already proposed, it handles these features separately, On the other hand, the proposed algorithm utilizes these features simultaneously to save computation time all the more. Since the proposed algorithm rejects those codewords that are impossible to be the nearest codeword, it produces the same output as the conventional full search algorithm. The simulation results confirm the effectiveness of the proposed algorithm.
TL;DR: A Bayesian solution to the problem of multiple direction of arrival detection/estimation from a single temporal observation of array data, where the array can be of general configuration, yields a novel DOA estimation cost function and a new detection algorithm useful for the single-snapshot case.
Abstract: We present a Bayesian solution to the problem of multiple direction of arrival (DOA) detection/estimation from a single temporal observation of array data, where the array can be of general configuration. The proposed method consists of multiple-step likelihood marginalizations using noninformative priors selected specifically to allow valid comparison of alternative model orders. This yields a novel DOA estimation cost function and a new detection algorithm useful for the single-snapshot case.
TL;DR: The algorithm uses a two-dimensional ESPRIT-like shift-invariance technique to separate and estimate the phase shifts due to delay and direction-of-incidence, with automatic pairing of the two parameter sets.
Abstract: Assuming a multipath propagation scenario, we derive a closed-form subspace-based method for the simultaneous estimation of arrival angles and path delays from measured channel impulse responses, using knowledge of the transmitted pulse shape function and assuming a uniform linear array and uniform sampling. The algorithm uses a two-dimensional (2-D) ESPRIT-like shift-invariance technique to separate and estimate the phase shifts due to delay and direction-of-incidence, with automatic pairing of the two parameter sets. A straightforward extension to the multiuser case allows to connect rays to users as well.
TL;DR: A new method for detecting microcalcifications in mammograms is described, which utilizes skewness and kurtosis as measures of the asymmetry and impulsiveness of the distribution.
Abstract: A new method for detecting microcalcifications in mammograms is described. In this method, the mammogram image is first processed by a subband decomposition filterbank. The bandpass subimage is divided into overlapping square regions in which skewness and kurtosis as measures of the asymmetry and impulsiveness of the distribution are estimated. The detection method utilizes these two parameters. A region with high positive skewness and kurtosis is marked as a region of interest. Simulation results show that this method is successful in detecting regions with microcalcifications.
TL;DR: New approaches to improve sparse application-specific language models by combining domain dependent and out-of-domain data are investigated, including a back-off scheme that effectively leads to context-dependent multiple interpolation weights, and a likelihood-based similarity weighting scheme to discriminatively use data to train a task- specific language model.
Abstract: Standard statistical language modeling techniques suffer from sparse data problems when applied to real tasks in speech recognition, where large amounts of domain-dependent text are not available. We investigate new approaches to improve sparse application-specific language models by combining domain dependent and out-of-domain data, including a back-off scheme that effectively leads to context-dependent multiple interpolation weights, and a likelihood-based similarity weighting scheme to discriminatively use data to train a task-specific language model. Experiments with both approaches on a spontaneous speech recognition task (switchboard), lead to reduced word error rate over a domain-specific n-gram language model, giving a larger gain than that obtained with previous brute-force data combination approaches.
TL;DR: Interpolated root-MUSIC with a virtual uniform linear array (ULA) of length x/sub n/-x/sub 1/ has better asymptotic performance than conventional root- MUSIC applied to a real ULA of the same length.
Abstract: Given an n-element linear array with the fixed positions x/sub 1/ and x/sub n/ of the leftmost and rightmost array sensors, it is shown that the stochastic Cramer-Rao bound (CRB) and MUSIC performance depend on positions of the remaining n-2 sensors within the interval [x/sub 1/, x/sub n/]. The asymptotic performance of the interpolated array approach shows similar dependence. The most favorable geometries are unrealizable for q
TL;DR: A new embedded zerotree wavelet image coding algorithm that is based on the algorithms developed by Shapiro and Said and is substantially more robust with respect to varying channel error conditions, which provides much needed reliability in low-bandwidth wireless applications.
Abstract: We present a new embedded zerotree wavelet image coding algorithm that is based on the algorithms developed by Shapiro (see IEEE Trans. Signal Processing, Spec. Issue Wavelets Signal Processing, vol.41, no.12, p.3445-62, 1993) and Said et al. (see IEEE Trans. Circuits Syst. Video Technol., vol.6, no.6, p.243-50, 1996). Our algorithm features a relatively simple coding structure and provides a better framework for balancing between high compression performance and robustness to channel errors. The fundamental approach is to explicitly classify the encoder's output bit sequence into subsequences, which are then protected differently according to their importance and robustness. Experimental results indicate that, for noisy channels, the proposed algorithm is slightly more resilient to channel errors than more complex and sophisticated source-channel coding algorithms. More important is that our algorithm is substantially more robust with respect to varying channel error conditions. This provides much needed reliability in low-bandwidth wireless applications.
TL;DR: The novel ML block matching techniques correspond to accurate statistical descriptions of ultrasound images, and are evaluated experimentally using sequences of transesophageal ultrasound images of the heart.
Abstract: Maximum likelihood (ML) techniques are defined for optimum block matching to enable motion estimation sequences of ultrasound B-mode images. Such motion estimation is needed as a diagnostic tool in medical use of ultrasound imagery. It is also needed for the efficient compression of sequences of ultrasound images. The novel ML block matching techniques correspond to accurate statistical descriptions of ultrasound images, and are evaluated experimentally using sequences of transesophageal ultrasound images of the heart.
TL;DR: A new method for the estimation of the signal subspace and noise subspace is introduced based on a joint block-diagonalization (JBD) of a combined set of spatio-temporal correlation matrices.
Abstract: Direction of arrival (DOA) estimation techniques require knowledge of the sensor-to-sensor correlation of the noise, which constitutes a significant drawback. In the case of temporally correlated signals, it is possible to estimate the signal parameters without any assumptions made on the spatial covariance matrix of the noise. A new method for the estimation of the signal subspace and noise subspace is introduced. The proposed approach is based on a joint block-diagonalization (JBD) of a combined set of spatio-temporal correlation matrices. Once the signal and the noise subspaces are estimated, any subspace based approach can be applied for DOA estimation. A performance comparison of the proposed approach with an existing technique is provided.
TL;DR: The resulting adaptive STFT shares many desirable properties with the adaptive CKD, such as the ability to adapt to transient as well as long-term signal components, making it competitive in complexity with nonadaptive time-frequency algorithms.
Abstract: This article presents a method of adaptively adjusting the window length used in short-time Fourier analysis, related to our earlier work in which we developed a means of adaptively optimizing the performance of the cone kernel distribution (CKD). The optimal CKD cone length is, by definition, a measure of the interval over which the signal has constant or slowly changing frequency structure. The article shows that this length can also be used to compute a time-varying short-time Fourier transform (STFT). The resulting adaptive STFT shares many desirable properties with the adaptive CKD, such as the ability to adapt to transient as well as long-term signal components. The optimization requires O(N) operations per step, less than the fast Fourier transform (FFT) used in computing each time slice, making it competitive in complexity with nonadaptive time-frequency algorithms.
TL;DR: This work addresses the problem of blind second-order equalization of scalar-valued polynomial channels and shows that the estimate is consistent, and this property remains unchanged in the presence of stationary noise and when the symbols are colored.
Abstract: We address the problem of blind second-order equalization of scalar-valued polynomial channels. When an almost periodic deterministic function modulates the initial symbol sequence, the observation exhibits second-order cyclostationarity. This property is shown to give rise to a structured spectral factorization problem. The identification of the unknown channel is then always possible via a structured version of the subspace method. No assumption on the channel is required, except the knowledge of a lower bound of the order. The estimate is consistent, and this property remains unchanged in the presence of stationary noise and when the symbols are colored.
TL;DR: A simple interleaving technique is presented that significantly improves the performance of subspace based methods in the case of closely spaced frequencies and is comparable to the corresponding Cramer-Rao bound (CRB).
Abstract: Both experience and analysis show that the widely used subspace methods, such as MUSIC and ESPRIT, perform poorly when applied to estimate closely spaced sinusoidal frequencies. This severely limits the resolution of subspace methods for frequency estimation. In this letter, we present a simple interleaving technique that significantly improves the performance of subspace based methods in the case of closely spaced frequencies. Simulation results show that the improved performance is comparable to the corresponding Cramer-Rao bound (CRB).
TL;DR: A variable stepsize is proposed that is based on the detection of speakers' activities, and involves estimates of system parameters and background noise, and real-time experiments demonstrate the workability in practice.
Abstract: Acoustic echo cancellation algorithms need a reliable adaptation control to cope with changing systems and nonstationary signals. A variable stepsize is proposed that is based on the detection of speakers' activities, and involves estimates of system parameters and background noise. Real-time experiments demonstrate the workability in practice.
TL;DR: The proposed method can significantly reduce coding artifacts of low bit-rate coded images, and at the same time guarantee that the resulting images satisfies the quantization error constraint.
Abstract: This letter presents a new approach to reduce coding artifacts in transform image coding We approach the problem in an estimation of each transform coefficient from its quantized version with its local mean and variance The proposed method can significantly reduce coding artifacts of low bit-rate coded images, and at the same time guarantee that the resulting images satisfies the quantization error constraint
TL;DR: The methods of penalized least-squares and cross-validation balance the bias-variance tradeoff and lead to a closed form expression for the estimator, which is simultaneously optimal in a "small-sample", predictive sum of squares sense and asymptotically optimal in the mean square sense.
Abstract: This letter develops an optimal, nonlinear estimator of a deterministic signal in noise. The methods of penalized least-squares and cross-validation (CV) balance the bias-variance tradeoff and lead to a closed form expression for the estimator. The estimator is simultaneously optimal in a "small-sample", predictive sum of squares sense and asymptotically optimal in the mean square sense.