[ Exchanging Faces in Images ]
Computer Graphics      
  Exchanging Faces in Images  
Original Target Image
  Pasting somebody's face into an existing image with traditional photo retouching and digital image processing tools has only been possible if both images show the face from the same viewpoint and with the same illumination. However, this is rarely the case for given pairs of images.  
  Our system exchanges faces across large differences in viewpoint and illumination. It is based on an algorithm that estimates 3D shape and texture of faces along with all relevant scene parameters, such as pose and lighting, from single images. Manual interaction is reduced to clicking on a set of about 7 feature points, and marking the hairline in the target image. In order to draw a face from a source image into a target image, we apply the 3D reconstruction algorithm to both. Then, the face from the source image is rendered with the parameters from the target image. Finally, the face is composited with a background layer, which is essentially the target image, and a foreground hair layer that contains strands of hair covering the face.  
  The system can be used for image processing, virtual try-on of hairstyles, and face recognition. By separating face identity from imaging conditions, our approach provides an abstract representation of images and a novel, high-level tool for image manipulation.  
Exchanging Faces in Images, Blanz, V., Scherbaum, K., Vetter, T. and Seidel, H. P.
In: Cani, M.-P. and Slater, M. (Eds) Proceedings of EUROGRAPHICS 2004,
Grenoble, France, 2004. pdf EG2004 Best Paper Award (1st Prize)

© Sascha Nesch 2009