By Topic

Vision, Image and Signal Processing, IEE Proceedings -

Issue 5 • Date Oct 1994

Filter Results

Displaying Results 1 - 11 of 11
  • Triple correlation analysis of binary sequences for codeword detection

    Page(s): 297 - 302
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (388 KB)  

    It is well known that multiple correlations of deterministic signals allow information about those signals to be extracted. The paper describes the application of triple correlation techniques to identify characteristics of several binary cyclic sequences commonly used in communications systems, especially those used in spread-spectrum systems. It is shown that the feedback function for a linear feedback shift register which generates an m-sequence can be found through the use of triple correlation. Examples of how to identify the feedback function when noise and data are present on the spread-spectrum signal are given View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Comparison of known signal detection schemes under a weakly dependent noise model

    Page(s): 303 - 310
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (432 KB)  

    The authors consider the discrete-time signal detection problem under the presence of additive noise exhibiting weak dependence. They first propose a weakly dependent noise model, in which the additive noise is modelled as a moving average process. They derive the locally optimum, memoryless, and one-memory detector test statistics under the model. The asymptotic performance of the one-memory detector is compared with that of the locally optimum and memoryless detectors. Specific examples for the asymptotic performance comparison of these detectors are considered. The authors also investigate the finite sample-size performance of several detectors through Monte-Carlo simulation. It is observed that the one-memory detector can achieve almost optimum performance at the expense of only one memory unit under the weakly dependent noise model, and is rather insensitive to slight model change View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Hybrid approach to speech recognition using hidden Markov models and Markov chains

    Page(s): 273 - 279
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (540 KB)  

    The paper presents a hybrid of a hidden Markov model and a Markov chain model for speech recognition. In this hybrid, the hidden Markov model is concerned with the time-varying property of spectral features, while the Markov chain accounts for the interdependence of spectral features. The log-likelihood scores of the two models, with respect to a given utterance, are combined by a postprocessor to yield a combined log-likelihood score for word classification. Experiments on speaker-independent and multispeaker isolated English alphabet recognition show that the hybrid outperformed both the hidden Markov model and the Markov chain model in terms of recognition View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Fast time-series adaptive-filtering algorithm based on the QRD inverse-updates method

    Page(s): 325 - 333
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (492 KB)  

    A new adaptive filtering algorithm for time-series data based on the QRD inverse updates method of Pan and Plemmons (1989) is derived from first principles. In common with other fast algorithms for time-series adaptive filtering, this algorithm only requires O(p) operations for the solution of a pth-order problem. Unlike previous fast algorithms based on the QRD technique, the algorithm presented here explicitly produces the transversal filter weights. Furthermore the derivation of the algorithm is achieved, quite simply, by means of signal-flow-graph manipulation. The relationship between this fast QRD inverse updates algorithm and the FTF algorithm is briefly discussed. The results of some preliminary computer simulations of the algorithm, using finite-precision floating-point arithmetic, are presented View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Multiband linear predictive speech coding at very low bit rates

    Page(s): 289 - 296
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (572 KB)  

    The authors describe the multiband linear predictive (MB-LPC) vocoder and its operation at 2.4 kb/s and 1.2 kb/s. The MB-LPC vocoder uses mixed excitation and exploits the advantages of both time and frequency domain speech coding techniques to produce natural sounding, good quality speech. Subjective performance of speech at 2.4 kb/s produced by the MB-LPC is very close to that for the 4.15 kb/s INMARSAT-M IMBE speech coder. Informal listening tests have shown that in most cases people could not tell the difference between the new 2.4 kb/s MB-LPC coder and the 4.15 kb/s INMARSAT-M IMBE coder View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Two-dimensional pattern analysis and classification using a complex orthogonal estimation algorithm

    Page(s): 339 - 347
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (540 KB)  

    An orthogonal estimation algorithm is derived for the estimation of parameters associated with the complex autoregressive boundary model of two-dimensional shapes. It is shown that the coefficients of the orthogonal model and the original system parameters are invariant to rotation around the origin, to the choice of starting point in tracing the boundary and to scale and translation. The error reduction ratios derived from the estimation algorithm can be used to detect which terms should be included in the system model, and classification based on the orthogonal parameters is shown to be less susceptible to incorrect model order specification. Classification based on orthogonal data sets is also derived, and it is demonstrated that this approach can avoid some problems associated with classification based on the model parameters alone View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Comparison of some noise-compensation methods for speech recognition in adverse environments

    Page(s): 280 - 288
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (600 KB)  

    A comparative study is presented of three noise-compensation schemes, namely spectral subtraction, Wiener filters, and noise adaptation, for hidden-Markov-model-based speech recognition in adverse environments. The noise-compensation methods are evaluated on a spoken-digit database, in the presence of car noise and helicopter noise at different signal-to-noise ratios. Experimental results demonstrate that the noise-compensation methods achieve a substantial improvement in recognition accuracy across a wide range of signal-to-noise ratios. At a signal-to-noise ratio of -6 dB the recognition accuracy is improved from 11% to 83%. The use of cepstral-time matrices as an improved speech representation is also considered, and their combination with the noise-compensation methods is shown. Experiments show that the cepstral-time matrix is a more robust feature than a vector of identical size, composed of a combination of cepstral and differential cepstral features View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Design of digital FIR notch filters

    Page(s): 334 - 338
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (256 KB)  

    A new procedure for designing digital FIR notch filters for a specified notch frequency ωd and 3-dB rejection bandwidth Δω has been proposed. Design formulae for computing the required weights and the length N for the filter have been derived. Illustrative examples confirming the new approach are also given View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Fractal model of a one-dimensional discrete signal and its implementation

    Page(s): 318 - 324
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (444 KB)  

    An extended iterated-function-system (IFS) interpolation method is presented for modelling for a given discrete signal. To reduce the computing complexity a suboptimal search algorithm with a robust technique for estimating the IFS affine-map parameters is introduced. Simulation results show that the IFS approach achieves a higher signal-to-noise ratio than does an existing approach based on autoregressive modelling View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Hermite sieve as a wavelet-like array for 1D and 2D signal decomposition

    Page(s): 348 - 356
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (708 KB)  

    A new class of an array of wavelet-like functions, derived from generalised Hermite polynomials and controlled by a scale parameter, has been used to create a multilayered representation for one- and two-dimensional signals. This representation, which is explicitly in terms of an array of coefficients, reminiscent of Fourier series, is stable. Among its other properties, (a) the shape of the resolution cell in the `phase-space' is variable even at a specified scale, depending on the nature of the signal under consideration; and (b) zero crossings at the various scales can be extracted directly from the coefficients. The new representation is illustrated by examples. However, there do remain some basic problems with respect to the new representation View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Variable-step-size LMS algorithm: new developments and experiments

    Page(s): 311 - 317
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (444 KB)  

    The variable-step-size least-mean-square (VSLMS) algorithm is explored and adopted for tracking of time-varying environments. Two implementations of the VSLMS algorithm are proposed. The emphasis is on implementations sizes with different step sizes at various taps of the adaptive filter. General analysis of the VSLMS algorithm appears to be somewhat involved. However, for one implementation a limited analysis of the algorithm is found possible. For this implementation it is shown that, when the input samples to the adaptive filter are zero-mean, Gaussian and uncorrelated with one another, the VSLMS algorithm can adapt itself to select the optimum set of step sizes which results in the best-tracking performance. Simulation experiments with the VSLMS algorithm show that, under fairly mild conditions, both of the proposed implementations adapt toward the optimum step sizes View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.