Examinando por Materia "Algorithm parameters"

Mostrando 1 - 2 de 2
  • No hay miniatura disponible
    Ítem
    Performance study of an admission controller for wireless networks
    (IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2014-06-01) Giraldo, D.; Jaramillo, J. J.; Universidad EAFIT. Departamento de Ciencias Básicas; Electromagnetismo Aplicado (Gema)
    Determining if there is enough bandwidth to allow admit a new flow into a wireless network is a complex task. To solve this problem, various methods have been proposed for estimating bandwidth. We study through simulations the performance of a new algorithm that has theoretical guarantees to correctly determine the available bandwidth without disturbing other flows that are currently being served. Specifically, we studied how the selection of the algorithm parameters affect its speed of convergence and accuracy of the estimates of available bandwidth. The results confirm the theoretical results and are used to determine the best parameters for rapid convergence with minimal estimation error. © 2012 IEEE.
  • No hay miniatura disponible
    Ítem
    Statistical tuning of adaptive-weight depth map algorithm
    (SPRINGER, 2011-01-01) Hoyos, Alejandro; Congote, John; Barandiaran, Inigo; Acosta, Diego; Ruiz, Oscar; Universidad EAFIT. Departamento de Ingeniería de Procesos; Desarrollo y Diseño de Procesos
    In depth map generation, the settings of the algorithm parameters to yield an accurate disparity estimation are usually chosen empirically or based on unplanned experiments. A systematic statistical approach including classical and exploratory data analyses on over 14000 images to measure the relative influence of the parameters allows their tuning based on the number of bad-pixels. Our approach is systematic in the sense that the heuristics used for parameter tuning are supported by formal statistical methods. The implemented methodology improves the performance of dense depth map algorithms. As a result of the statistical based tuning, the algorithm improves from 16.78% to 14.48% bad-pixels rising 7 spots as per the Middlebury Stereo Evaluation Ranking Table. The performance is measured based on the distance of the algorithm results vs. the Ground Truth by Middlebury. Future work aims to achieve the tuning by using significantly smaller data sets on fractional factorial and surface-response designs of experiments. © 2011 Springer-Verlag.