<> "The repository administrator has not yet configured an RDF license."^^ . <> . . . "Efficient illumination independent appearance-based face tracking"^^ . "One of the major challenges that visual tracking algorithms face nowadays is being\r\nable to cope with changes in the appearance of the target during tracking. Linear\r\nsubspace models have been extensively studied and are possibly the most popular\r\nway of modelling target appearance. We introduce a linear subspace representation\r\nin which the appearance of a face is represented by the addition of two approxi-\r\nmately independent linear subspaces modelling facial expressions and illumination\r\nrespectively. This model is more compact than previous bilinear or multilinear ap-\r\nproaches. The independence assumption notably simplifies system training. We only\r\nrequire two image sequences. One facial expression is subject to all possible illumina-\r\ntions in one sequence and the face adopts all facial expressions under one particular\r\nillumination in the other. This simple model enables us to train the system with\r\nno manual intervention. We also revisit the problem of efficiently fitting a linear\r\nsubspace-based model to a target image and introduce an additive procedure for\r\nsolving this problem. We prove that Matthews and Baker’s Inverse Compositional\r\nApproach makes a smoothness assumption on the subspace basis that is equiva-\r\nlent to Hager and Belhumeur’s, which worsens convergence. Our approach differs\r\nfrom Hager and Belhumeur’s additive and Matthews and Baker’s compositional ap-\r\nproaches in that we make no smoothness assumptions on the subspace basis. In the\r\nexperiments conducted we show that the model introduced accurately represents\r\nthe appearance variations caused by illumination changes and facial expressions.\r\nWe also verify experimentally that our fitting procedure is more accurate and has\r\nbetter convergence rate than the other related approaches, albeit at the expense of\r\na slight increase in computational cost. Our approach can be used for tracking a\r\nhuman face at standard video frame rates on an average personal computer."^^ . "2009-04-02" . . . "27" . "5" . . "Image and Vision Computing"^^ . . . "02628856" . . . . . . . . . . . . . "José Miguel"^^ . "Buenaposada Biencinto"^^ . "José Miguel Buenaposada Biencinto"^^ . . "Luis"^^ . "Baumela Molina"^^ . "Luis Baumela Molina"^^ . . "Enrique"^^ . "Muñoz"^^ . "Enrique Muñoz"^^ . . . . . . "Efficient illumination independent appearance-based face tracking (PDF)"^^ . . . "BUENAPO_ART2009_01.pdf"^^ . . . "Efficient illumination independent appearance-based face tracking (Other)"^^ . . . . . . "indexcodes.txt"^^ . . . "Efficient illumination independent appearance-based face tracking (Other)"^^ . . . . . . "small.jpg"^^ . . . "Efficient illumination independent appearance-based face tracking (Other)"^^ . . . . . . "medium.jpg"^^ . . . "Efficient illumination independent appearance-based face tracking (Other)"^^ . . . . . . "preview.jpg"^^ . . . "Efficient illumination independent appearance-based face tracking (Other)"^^ . . . . . . "lightbox.jpg"^^ . . "HTML Summary of #1861 \n\nEfficient illumination independent appearance-based face tracking\n\n" . "text/html" . . . "Matemáticas"@es . "Mathematics"@en . . . "Informática"@es . "Computer Science"@en . .