作者: Thomas Vetter , Nikolaus Troje
DOI: 10.1007/978-3-642-79980-8_14
关键词: Representation (mathematics) 、 Artificial intelligence 、 Small data 、 Principal axis theorem 、 Facial recognition system 、 Computer vision 、 Face (geometry) 、 Separation (aeronautics) 、 Set (abstract data type) 、 Texture (geology) 、 Computer science
摘要: Human faces differ in shape and texture. This paper describes a representation of grey-level images human based on an automated separation two- dimensional The was done using the point correspondence between different images, which established through algorithms known from optical flow computation. A linear description separated texture spaces allows smooth modeling faces. Pictures along principal axes small data set 50 are shown. We also show face reconstructions this example set.