publications:
- title: "Error concealment by means of motion refinement and regularized bregman divergence"
  author:
  - name: "Alessandra Martins Coelho"
    link: "http://academic.research.microsoft.com/Author/53930956/alessandra-martins-coelho"
  - name: "Vania Vieira Estrela"
    link: "http://academic.research.microsoft.com/Author/3361386/vania-v-estrela"
  - name: "Felipe Pereira do Carmo"
    link: "http://academic.research.microsoft.com/Author/49826901/f-p-do-carmo"
  - name: "Sandro Roberto Fernandes"
    link: "http://academic.research.microsoft.com/Author/53603335/sandro-r-fernandes"
  - name: "Vania Vieira Estrela"
    link: "http://academic.research.microsoft.com/Publication/584912/regularized-pel-recursive-motion-estimation-using-generalized-cross-validation-and-spatial"
  - name: "Vania Vieira Estrela"
    link: "http://academic.research.microsoft.com/Publication/584912/regularized-pel-recursive-motion-estimation-using-generalized-cross-validation-and-spatial"
  - name: "Vania Vieira Estrela"
    link: "https://www.researchgate.net/profile/Vania_Estrela?ev=hdr_xprf&_sg=l2it7UJUf2g4wSUGGVaee_gK6wbXK75hN_Rt6P_99NMr2_HObKqAkOwo4ujJMZp7"
  - name: "Alessandra Martins Coelho"
    link: "http://academic.research.microsoft.com/Author/53930956/alessandra-martins-coelho"
  - name: "Sandro Roberto Fernandes"
    link: "http://academic.research.microsoft.com/Author/53603335/sandro-r-fernandes"
  - name: "Alessandra Martins Coelho"
    link: "http://academic.research.microsoft.com/Author/53930956/alessandra-martins-coelho"
  - name: "Sandro Roberto Fernandes"
    link: "http://academic.research.microsoft.com/Author/53603335/sandro-r-fernandes"
  - name: "Alessandra Martins Coelho"
    link: "http://academic.research.microsoft.com/Author/53930956/alessandra-martins-coelho"
  - name: "Alessandra Martins Coelho"
    link: "http://academic.research.microsoft.com/Author/53930956/alessandra-martins-coelho"
  - name: "Vânia Estrela"
    link: "http://br.linkedin.com/pub/vania-v-estrela/29/9bb/96b/"
  - name: "Vânia Vieira Estrela"
    link: "http://scholar.google.com/citations?hl=en&user=-ZbglqgAAAAJ&view_op=list_works&pagesize=100"
  - name: "Vania Vieira Estrela"
    link: "https://www.researchgate.net/profile/Vania_Estrela?ev=hdr_xprf&_sg=l2it7UJUf2g4wSUGGVaee_gK6wbXK75hN_Rt6P_99NMr2_HObKqAkOwo4ujJMZp7"
  year: "2012"
  abstract: "This work addresses the problem of error concealment in video transmission systems over noisy channels employing Bregman divergences along with regularization. Error concealment intends to improve the effects of disturbances at the reception due to bit-errors or cell loss in packet networks. Bregman regularization gives accurate answers after just some iterations with fast convergence, better accuracy and stability. This technique has an adaptive nature: the regularization functional is updated according to Bregman functions that change from iteration to iteration according to the nature of the neighborhood under study at iteration n. Numerical experiments show that high-quality regularization parameter estimates can be obtained. The convergence is sped up while turning the regularization parameter estimation less empiric, and more automatic."
  links:
    published: "https://researchr.org/publication/citeulike%3A11098377"
  researchr: "https://researchr.org/publication/preprint-citeulike%3A11098377"
  cites: 0
  citedby: 0
  type: "Preprint"
  kind: "techreport"
  key: "preprint-citeulike:11098377"
- title: "Error concealment by means of motion refinement and regularized bregman divergence"
  author:
  - name: "Alessandra Martins Coelho"
    link: "http://academic.research.microsoft.com/Author/53930956/alessandra-martins-coelho"
  - name: "Vania Vieira Estrela"
    link: "http://academic.research.microsoft.com/Author/3361386/vania-v-estrela"
  - name: "Felipe Pereira do Carmo"
    link: "http://academic.research.microsoft.com/Author/49826901/f-p-do-carmo"
  - name: "Sandro Roberto Fernandes"
    link: "http://academic.research.microsoft.com/Author/53603335/sandro-r-fernandes"
  - name: "Vania Vieira Estrela"
    link: "http://academic.research.microsoft.com/Publication/584912/regularized-pel-recursive-motion-estimation-using-generalized-cross-validation-and-spatial"
  - name: "Vania Vieira Estrela"
    link: "http://academic.research.microsoft.com/Publication/584912/regularized-pel-recursive-motion-estimation-using-generalized-cross-validation-and-spatial"
  - name: "Vania Vieira Estrela"
    link: "https://www.researchgate.net/profile/Vania_Estrela?ev=hdr_xprf&_sg=l2it7UJUf2g4wSUGGVaee_gK6wbXK75hN_Rt6P_99NMr2_HObKqAkOwo4ujJMZp7"
  - name: "Alessandra Martins Coelho"
    link: "http://academic.research.microsoft.com/Author/53930956/alessandra-martins-coelho"
  - name: "Sandro Roberto Fernandes"
    link: "http://academic.research.microsoft.com/Author/53603335/sandro-r-fernandes"
  - name: "Alessandra Martins Coelho"
    link: "http://academic.research.microsoft.com/Author/53930956/alessandra-martins-coelho"
  - name: "Sandro Roberto Fernandes"
    link: "http://academic.research.microsoft.com/Author/53603335/sandro-r-fernandes"
  - name: "Alessandra Martins Coelho"
    link: "http://academic.research.microsoft.com/Author/53930956/alessandra-martins-coelho"
  - name: "Alessandra Martins Coelho"
    link: "http://academic.research.microsoft.com/Author/53930956/alessandra-martins-coelho"
  - name: "Vânia Estrela"
    link: "http://br.linkedin.com/pub/vania-v-estrela/29/9bb/96b/"
  - name: "Vânia Vieira Estrela"
    link: "http://scholar.google.com/citations?hl=en&user=-ZbglqgAAAAJ&view_op=list_works&pagesize=100"
  - name: "Vania Vieira Estrela"
    link: "https://www.researchgate.net/profile/Vania_Estrela?ev=hdr_xprf&_sg=l2it7UJUf2g4wSUGGVaee_gK6wbXK75hN_Rt6P_99NMr2_HObKqAkOwo4ujJMZp7"
  year: "2012"
  doi: "10.1007/978-3-642-32639-4_78"
  abstract: "This work addresses the problem of error concealment in video transmission systems over noisy channels employing Bregman divergences along with regularization. Error concealment intends to improve the effects of disturbances at the reception due to bit-errors or cell loss in packet networks. Bregman regularization gives accurate answers after just some iterations with fast convergence, better accuracy and stability. This technique has an adaptive nature: the regularization functional is updated according to Bregman functions that change from iteration to iteration according to the nature of the neighborhood under study at iteration n. Numerical experiments show that high-quality regularization parameter estimates can be obtained. The convergence is sped up while turning the regularization parameter estimation less empiric, and more automatic."
  links:
    technicalreport: "https://researchr.org/publication/preprint-citeulike%3A11098377"
    "url": "http://dx.doi.org/10.1007/978-3-642-32639-4\\_78"
  tags:
  - "empirical"
  - "optical flow"
  - "regularization"
  - "refinement"
  - "monitoring"
  - "clustering"
  - "Inverse problems"
  - "image processing"
  - "Computer vision"
  - "Motion estimation"
  - "surveillance"
  - " motion detection"
  - "Error concealment"
  - "Bregman distance"
  researchr: "https://researchr.org/publication/citeulike%3A11098377"
  cites: 16
  citedby: 4
  pages: "650-657"
  booktitle: "Proceedings of the 13th international conference on Intelligent Data Engineering and Automated Learning"
  volume: "7435"
  series: "IDEAL'12"
  address: "Berlin, Heidelberg"
  publisher: "Springer-Verlag"
  isbn: "978-3-642-32638-7"
  kind: "inproceedings"
  key: "citeulike:11098377"
- title: "EM-Based Mixture Models Applied to Video Event Detection"
  author:
  - name: "Alessandra Martins Coelho"
    link: "http://academic.research.microsoft.com/Author/53930956/alessandra-martins-coelho"
  - name: "Vania Vieira Estrela"
    link: "http://buscatextual.cnpq.br/buscatextual/visualizacv.do?id=K4793789Y3"
  - name: "Vania Vieira Estrela"
    link: "http://academic.research.microsoft.com/Publication/584912/regularized-pel-recursive-motion-estimation-using-generalized-cross-validation-and-spatial"
  - name: "Vania Vieira Estrela"
    link: "http://academic.research.microsoft.com/Publication/584912/regularized-pel-recursive-motion-estimation-using-generalized-cross-validation-and-spatial"
  - name: "Vania Vieira Estrela"
    link: "http://academic.research.microsoft.com/Publication/584912/regularized-pel-recursive-motion-estimation-using-generalized-cross-validation-and-spatial"
  - name: "Vania Vieira Estrela"
    link: "https://www.researchgate.net/profile/Vania_Estrela?ev=hdr_xprf&_sg=l2it7UJUf2g4wSUGGVaee_gK6wbXK75hN_Rt6P_99NMr2_HObKqAkOwo4ujJMZp7"
  - name: "Vania Vieira Estrela"
    link: "https://www.researchgate.net/profile/Vania_Estrela?ev=hdr_xprf&_sg=l2it7UJUf2g4wSUGGVaee_gK6wbXK75hN_Rt6P_99NMr2_HObKqAkOwo4ujJMZp7"
  - name: "Alessandra Martins Coelho"
    link: "http://academic.research.microsoft.com/Author/53930956/alessandra-martins-coelho"
  year: "2012"
  abstract: "Surveillance system (SS) development requires hi-tech support to prevail over the shortcomings related to the massive quantity of visual information from SSs. Anything but reduced human monitoring became impossible by means of its physical and economic implications, and an advance towards an automated surveillance becomes the only way out. When it comes to a computer vision system, automatic video event comprehension is a challenging task due to motion clutter, event understanding under complex scenes, multi-level semantic event inference, contextualization of events and views obtained from multiple cameras, unevenness of motion scales, shape changes, occlusions and object interactions among lots of other impairments. In recent years, state-of-the-art models for video event classification and recognition  include modeling events to discern context, detecting incidents with only one camera, low-level feature extraction and description, high-level semantic event classification and recognition. Even so, it is still very burdensome to recuperate or label a specific video part relying solely on its content. Principal component analysis (PCA) has been widely known and used, but when combined with other techniques such as the expectation-maximization (EM) algorithm its computation becomes more efficient. This chapter introduces advances associated to the concept of Probabilistic PCA (PPCA) analysis [Tipping et al., 1999)] by of video event understanding technologies. The PPCA model-based method results from the combination of a linear model and the EM algorithm in an iterative fashion in order to determine a principal subspace (PS). Thus, additional work may be needed to find precise principal eigenvectors of the data covariance matrix, with no rotational uncertainty. Kernel principal component analysis (KPCA) is a nonlinear PCA extension that relies on the kernel trick. It has received immense consideration for its value in nonlinear feature mining and other applications. On the other hand, the main drawback of the standard KPCA is that the huge amount of computation required, and the space needed to store the kernel matrix. KPCA can be viewed as a primal space problem with samples created via incomplete Cholesky decomposition. Therefore, all the efficient PCA algorithms can be easily adapted into KPCA. Furthermore, KPCA can be extended to a mixture of local KPCA models by applying the mixture model to probabilistic PCA in the primal space. The theoretical analysis and experiments can shed light onthe performance of the proposed methods in terms of computational efficiency and storage space, as well as recognition rate, especially when the number of data points n is large. By considering KPCA from a probabilistic point of view with the help of the EM algorithm, the computational load can be alleviated, but there still exists a rotational ambiguity with the resulting algorithm implementation. To unravel this intricacy, a constrained EM algorithm for KPCA (and PCA) was formulated founded on a coupled probability model. This brings in advantages related to many factors such as the necessary precision of extracted components, the number of the separated smaller data sets (which is usually empirically set), and the data to be processed. As a generic methodology, another thread of speeding up kernel machine learning is to seek a low-rank approximation to the kernel matrix. Since, as noted by several researchers, the spectrum of the kernel matrix tends to decay rapidly, the low-rank approximation often achieves satisfactory precision. This chapter also aims at looking closely to ways and metrics in order to evaluate these less intensive EM implementations of PCA and KPCA. "
  tags:
  - "rule-based"
  - "Expectation-Maximization Algorithm"
  - "process monitoring"
  - "EM Algorithm"
  - "data-flow analysis"
  researchr: "https://researchr.org/publication/coelhoestrela2012a"
  cites: 22
  citedby: 11
  howpublished: "Printed and online"
  kind: "misc"
  key: "coelhoestrela2012a"