By Topic

Applied Statistical Pattern Recognition (Ref. No. 1999/063), IEE Colloquium on

Date 1999

Filter Results

Displaying Results 1 - 17 of 17
  • Support vector machines: a tutorial overview and critical appraisal

    Publication Year: 1999
    Cited by:  Papers (2)
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (72 KB)

    Summary form only given. There has been much interest in the use of support vector machines (SVM) as an approach to high performance pattern classification. In the linearly separable case, SVMs attempt to position a class boundary so that the margin from the nearest example is maximised. This criterion can be implemented by solving a quadratic programming problem, and the solution turns out to be ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Detecting cells in DIC microscope images using a high level Bayesian model and template matching

    Publication Year: 1999, Page(s):7/1 - 7/6
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (868 KB)

    Identification of objects in differential interference contrast (DIC) microscope images by digital image analysis is a hard task. While DIC microscopy is well suited to visualisation of near-transparent cells, the microscope optics cause a pattern of light and dark cell edges to appear in the image, giving a pseudo 3D effect. A high level Bayesian statistical approach is described as an alternativ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Non-linear feature space transformations

    Publication Year: 1999, Page(s):17/1 - 17/5
    Cited by:  Patents (1)
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (400 KB)

    Linear methods are strongly preferred in statistical pattern recognition, but problems in perception require nonlinear analysis and operators. Even the most successful linear methods lack robustness, especially when the normal variation in the data reveals new structure. An alternative to computing complex features or devising a complex decision rule is to transform the feature space so that the s... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Robust model-based signal analysis and identification

    Publication Year: 1999, Page(s):4/1 - 4/7
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (480 KB)

    We describe and evaluate a model-based scheme for feature extraction and model-based signal identification, which uses likelihood criteria for edge detection. Likelihood measures from the feature identification process are shown to provide a well behaved measure of signal interpretation confidence. We demonstrate that complex transient signals, from one of 6 classes, can reliably be identified at ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Minimum description length and the inference of scene structure from images

    Publication Year: 1999, Page(s):9/1 - 9/6
    Cited by:  Papers (1)
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (468 KB)

    Model selection is a central task in computer vision. The minimum description length (MDL) method links model selection to data compression: the best model is the one which yields the largest compression of the data. The general theoretical framework for compression is Kolmogorov complexity. MDL differs from Bayesian model selection (BMS) in that it is biased against complex probability density fu... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A little introduction to wavelets

    Publication Year: 1999, Page(s):1/1 - 1/6
    Cited by:  Papers (4)
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (352 KB)

    Wavelets have become a popular and useful tool for the investigation and analysis of many kinds of problem. It is difficult to pin down a unique origin for wavelet theory. Many fields provided the seeds from which wavelet theory grew. These include signal processing, physics and mathematics. The range of applications that wavelets are being applied to is large and growing. This paper reviews brief... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Using statistical models to interpret complex and variable images

    Publication Year: 1999, Page(s):6/1 - 6/4
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (320 KB)

    Model-based vision has been applied successfully to images of man-made objects. It has proved much more difficult to develop model-based approaches to interpreting images of complex and variable structures such as faces or the internal organs of the human body. The key problem is that of variability. Recent developments have shown that specific patterns of variability in shape and grey-level appea... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Analysis of the response of an auto adaptive matched filter

    Publication Year: 1999, Page(s):16/1 - 16/7
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (308 KB)

    This paper analyses the statistical response of an adaptive filter using quadratic form theory. Issues arising from the analysis are discussed and the results are compared to the those obtained using Monte Carlo simulations View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Recognizing anomalies in weather radar images

    Publication Year: 1999, Page(s):3/1 - 3/5
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (300 KB)

    The objective of a weather radar is to determine the locations and characteristics of meteorological objects, such as clouds, in the sky. Any echo coming back from a non-meteorological object (mountains, buildings, planes, birds, insects, etc.) is considered as being clutter. The delay between the times of transmission and return indicates the distance of the reflecting object from the radar. The ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Textured image segmentation using multiresolution Markov random fields

    Publication Year: 1999, Page(s):8/1 - 8/6
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (544 KB)

    A new stochastic image model, based on a multiresolution representation, is presented and used in the Bayesian segmentation of textured images. Results on a variety of images are used to demonstrate its effectiveness View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • IEE Colloquium on Applied Statistical Pattern Recognition (Ref. No. 1999/063)

    Publication Year: 1999
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (252 KB)

    The following topics were dealt with: statistical pattern recognition; wavelets; support vector machines; 2D shape recognition; signal separation; one-to-many mappings; radial basis function networks; auto-adaptive matched filter; nonlinear feature space transformations; weather radar image recognition; signal analysis; statistical image restoration; complex image interpretation; cell detection in... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A classification approach for estimating prior models in statistical image restoration

    Publication Year: 1999, Page(s):5/1 - 5/6
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (352 KB)

    We present a novel application of Bayesian classifier to the problem of statistical image restoration which has attracted much attention in recent years. Specifically, maximum a posteriori (MAP) based approaches have been applied fairly successfully to a variety of image data. An integral part of such restoration schemes is the estimation of the prior probability of occurrence of an estimate of th... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Texture recognition or classification using statistics

    Publication Year: 1999, Page(s):10/1 - 10/6
    Cited by:  Papers (1)
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (492 KB)

    Texture recognition/classification is one of the long standing problems of image processing. In this paper we regard this task as a `pattern recognition problem in the statistical parameter space'. There are three stages in this process - statistical feature extraction, statistical feature selection, and classification. We focus on the second and third order statistics for feature extraction. We u... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An introduction to robust shape classification

    Publication Year: 1999, Page(s):11/1 - 11/6
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (204 KB)

    Robust shape classifiers are compared, and it is found that conventional techniques based on the sample auto-covariance function suffer catastrophic reductions in performance in outlier contaminated data. However, robust procedures suffer much less degradation, with the robust spectral approach giving the best performance. The use of lag selection in the classification phase may be of independent ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Recognising 2-D shapes from incomplete boundaries

    Publication Year: 1999, Page(s):12/1 - 12/6
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (440 KB)

    In many cases where object identification is necessary complete boundaries are not available due to poor contrast or occlusion. In such cases global shape descriptors are not effective. We describe a multiresolution hypothesis and verify a method for matching characteristic local segments of a boundary with pre-defined models. Boundary shape is characterised using a model of the distribution of do... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • One-to-many mappings, continuity constraints and latent variable models

    Publication Year: 1999, Page(s):14/1 - 14/6
    Cited by:  Papers (1)
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (556 KB)

    We approach the problem of multivariate regression using latent variable models, which infer a low-dimensional representation of an observed, high-dimensional process. Defining functional relationships between variables may be conveniently done by picking informative points from the corresponding conditional distribution. However, this is problematic when this conditional distribution is multimoda... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Knowledge extraction and insertion from radial basis function networks

    Publication Year: 1999, Page(s):15/1 - 15/6
    Cited by:  Papers (3)
    IEEE is not the copyright holder of this material | Click to expandAbstract | PDF file iconPDF (424 KB)

    Neural networks provide excellent solutions for pattern recognition and classification problems. Unfortunately, in the case of distributed neural networks such as the multilayer perceptron it is difficult to comprehend the learned internal mappings. This makes any form of explanation facility such as that possessed by expert systems impractical. However, in the case of localist neural representati... View full abstract»

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