Artigo de evento

URI Permanente para esta coleção

https://repositorio.fei.edu.br/handle/FEI/5120

Navegar

Navegando Artigo de evento por Autor "Carlos E. Thomaz"

Agora exibindo 1 - 3 de 3
  • Nenhuma Miniatura disponível
    Artigo de evento
    A kernel maximum uncertainty discriminant analysis and its application to face recognition
    (2009-02-05) Carlos E. Thomaz; GIRALDI, G. A.
    In this paper, we extend the Maximum uncertainty Linear Discriminant Analysis (MLDA), proposed recently for limited sample size problems, to its kernel version. The new Kernel Maximum uncertainty Discriminant Analysis (KMDA) is a two-stage method composed of Kernel Principal Component Analysis (KPCA) followed by the standard MLDA. In order to evaluate its effectiveness, experiments on face recognition using the well-known ORL and FERET face databases were carried out and compared with other existing kernel discriminant methods, such as Generalized Discriminant Analysis (GDA) and Regularized Kernel Discriminant Analysis (RKDA). The classification results indicate that KMDA performs as well as GDA and RKDA, with the advantage of being a straightforward stabilization approach for the within-class scatter matrix that uses higher-order features for further classification improvements.
  • Nenhuma Miniatura disponível
    Artigo de evento
    A multivariate statistical analysis of muscular biopotencial for human arm movement characterization
    (2009-01-14) SILVA, G. A. DA; Castro, M.C.F.; Carlos E. Thomaz
    Pattern recognition of electromyographic signals consists of a hard task due to the high dimensionality of the data and noise presence on the acquired signals. This work intends to study the data set as a multivariate pattern recognition problem by applying linear transformations to reduce the data dimensionality. Five volunteers contributed in a previous experiment that acquired the myoelectrical signals using surface electrodes. Attempts to analyse the groups of acquired data by means of descriptive statistics have shown to be inconclusive. This works shows that the use of multivariate statistical techniques such as Principal Components Analysis (PCA) and Maximum uncertainty Linear Discriminant Analysis (MLDA) to characterize the: acquired set of signals through low dimensional scatter plots provides a new understanding of the data spread, making easier its analysis. Considering the arm horizontal movement and the acquired set of data used in this research, a multivariate linear separation between the patterns of interest quantified by the distance of Bhattacharyya suggests that it's possible not only to characterize the angular joint position, but also to confirm that different movements recruit similar amounts of energy to be executed.
  • Nenhuma Miniatura disponível
    Artigo de evento
    Low-cost automatic inpainting for artifact suppression in facial images
    (2013-02-21) SOBIECKI, A.; TELEA, A.; GIRALDI, G.; NEVES, L. A. P.; Carlos E. Thomaz
    Facial images are often used in applications that need to recognize or identify persons. Many existing facial recognition tools have limitations with respect to facial image quality attributes such as resolution, face position, and artifacts present in the image. In this paper we describe a new low-cost framework for preprocessing low-quality facial images in order to render them suitable for automatic recognition. For this, we first detect artifacts based on the statistical difference between the target image and a set of pre-processed images in the database. Next, we eliminate artifacts by an inpainting method which combines information from the target image and similar images in our database. Our method has low computational cost and is simple to implement, which makes it attractive for usage in low-budget environments. We illustrate our method on several images taken from public surveillance databases, and compare our results with existing inpainting techniques.