Recovering 3D Shape and Albedo from a Face Image under Arbitrary Lighting and Pose by Using a 3D Illumination-Based AAM Model

摘要: We present a novel iterative approach for recovering 3D shape and albedo from face images affected by non-uniform lighting non-frontal pose. fit active appearance model based on illumination, to image. In contrast other works where an initial pose is required, we only need simple initialization in translation scale. Our optimization method improves the Jacobian each iteration using parameters of estimated previous iterations. fitting algorithm obtains compact set albedo, shape, illumination which describe input show that our able accurately estimate are strongly related identity. Experimental results proposed can be easily extended recognition under variations.