作者: Thomas Vetter , Jean-Sebastian Pierrard , Sami Romdhani
DOI:
关键词: Radius of convergence 、 Artificial intelligence 、 Image (mathematics) 、 Object based 、 Object (computer science) 、 Face (geometry) 、 Texture (music) 、 Fitting algorithm 、 Geography 、 Computer vision 、 Pattern recognition 、 Improved algorithm
摘要: Disclosed is an improved algorithm for estimating the 3D shape of a 3-dimensional object, such as human face, based on information retrieved from single photograph by recovering parameters model and methods systems using same. Beside pixel intensity, invention uses various image features in multi-features fitting (MFF) that has wider radius convergence higher level precision provides thereby better results.