• Advertisement

hugomcp

Member
  • Content count

    3
  • Joined

  • Last visited

Community Reputation

122 Neutral

About hugomcp

  • Rank
    Newbie
  1. Dear all As you are probably aware, during 2007-2009 the SOCIA Lab. (Soft Computer And Image Analysis Group, University of Beira Interior, Portugal) held an international contest for iris biometric purposes. It focused on the segmentation and noise detection of iris images captured in the visible wavelength and under unconstrained imaging setups. This contest (NICE.I - Noisy Iris Challenge Evaluation, http://nice1.di.ubi.pt) received over 90 participations from over 20 different countries and the best 8 participants were invited to publish their method in a special issue of the Image and Vision Computing Journal (IVC, to appear in 2010). Now, we are organizing the complementary part of the contest (NICE:II) that will comprise the signatures encoding and matching of previously segmented noisy iris images. This will complete the evaluation of the most traditional stages of iris recognition systems: segmentation + noise detection (NICE.I, IVC Special Issue) and signatures encoding + matching (NICE:II, Pattern Recognition Letters Special Issue). Once again, the idea is to publish a special issue describing the best 8 to 10 methods of the contest, which hopefully will constitute an important step toward the development of less constrained iris recognition systems. We have already developed the software for the automatic evaluation of participations and the web page with the most important information is online (please, check more information at http://nice2.di.ubi.pt). Briefly, the NICE:II is a completely free-of-charge iris encoding and matching contest that operates on noisy data acquired under less constrained image capturing conditions and at the visible wavelength. As before stated, the participations that achieve the lowest error rates will be invited to publish their approach in the Pattern Recognition Letters Journal, (ISI Web-of-Knowledge indexed), in the NICE:II special issue. Detailed information can be found at the contest web site: http://nice2.di.ubi.pt and in the "pdf" file that is attached to this email. We look forward for your participation. You are also invited to propose the participation in the NICE:II to undergraduated, MSc. or PhD. students under your supervision. Best regards ------------------------------------------------------ Hugo Pedro Proenca Member of the NICE:II Organizing Committee, Department of Computer Science, University of Beira Interior, Covilha, Portugal. http://www.di.ubi.pt/~hugomcp
  2. Dear Srs As you are probably aware, during 2007-2009 the SOCIA Lab. (Soft Computer And Image Analysis Group, University of Beira Interior, Portugal) held an international contest for iris biometric purposes. It focused on the segmentation and noise detection of iris images captured in the visible wavelength and under unconstrained imaging setups. This contest (NICE.I - Noisy Iris Challenge Evaluation, http://nice1.di.ubi.pt) received over 90 participations from over 20 different countries and the best 8 participants were invited to publish their method in a special issue of the Image and Vision Computing Journal (IVC, to appear in 2010). Now, we are organizing the complementary part of the contest (NICE:II) that will comprise the signatures encoding and matching of previously segmented noisy iris images. This will complete the evaluation of the most traditional stages of iris recognition systems: segmentation + noise detection (NICE.I, IVC Special Issue) and signatures encoding + matching (NICE:II, Pattern Recognition Letters Special Issue). Once again, the idea is to publish a special issue describing the best 8 to 10 methods of the contest, which hopefully will constitute an important step toward the development of less constrained iris recognition systems. We have already developed the software for the automatic evaluation of the participations and the web page with the most important information is online (please, check more information at http://nice2.di.ubi.pt). Briefly, the NICE:II is a completely free-of-charge iris encoding and matching contest that operates on noisy data acquired under less constrained image conditions and in the visible wavelength. As before stated, the participations that achieve the lowest error rates will be invited to publish their approach in the Pattern Recognition Letters Journal, (ISI Web-of-Knowledge indexed), in the NICE:II special issue. Detailed information can be found at the contest web site: http://nicee.di.ubi.pt and in the "pdf" file that is attached to this email. We look forward for your participation. You are also invited to propose the participation in the NICE:II to undergraduated, MSc. or PhD. students under your supervision. Best regards ------------------------------------------------------ Hugo Pedro Proenca Member of the NICE:II Organizing Committee, Department of Computer Science, University of Beira Interior, Covilha, Portugal. http://www.di.ubi.pt/~hugomcp
  3. The SOCIA Lab. (Soft Computing and Image Analysis Group) of the Department of Computer Science, University of Beira Interior (Portugal) is organizing a new iris segmentation contest: NICE.I (Noisy Iris Challenge Evaluation - Part I). Within this laboratory we developed and publicly released the UBIRIS database (http://iris.di.ubi.pt). Since September, 2005, this database has been freely downloaded by more than 500 individuals and institutions from over 70 countries and its data constituted the basis for a large number of works with academic, research and commercial purposes. In the iris recognition compass, the robustness to noise receives growing attention. Many efforts are focused toward the accurate iris recognition within less constrained image capturing environments, probably toward the covert iris recognition in a near future. To support such research developments, we created a new version of the UBIRIS database (UBIRIS.v2). This version is significantly extended, has new and more realistic noise factors and will be used as data source for the Noisy Iris Challenge Contest - Part I (NICE.I). Basically, the NICE.I is a completely free-of-charge iris segmentation and noise detection contest that operates on noisy data, resultant from less constrained image capturing conditions. Its main purpose is the evaluation of the iris segmentation robustness to noise factors. Its is open to individuals and institutions, either with academic, research or commercial purposes. Moreover, the NICE.I contest will be the only way to legally obtain the UBIRIS.v2 database within the next 24 months. Also, the participants that achieve the lowest error rates on the NICE.I contest will be invited to publish their approach on the Elsevier Image and Vision Computing Journal, (ISI Web of Knowledge indexed) through a journal's special issue. Detailed information can be found at the contest web site: http://nice1.di.ubi.pt We look forward for your participation, Best regards Hugo Proença Member of the NICE.I Organizing Committee
  • Advertisement