Efficient, Robust and Accurate Fitting of a 3D Morphable Model

  • Authors:
  • Sami Romdhani;Thomas Vetter

  • Affiliations:
  • -;-

  • Venue:
  • ICCV '03 Proceedings of the Ninth IEEE International Conference on Computer Vision - Volume 2
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

3D Morphable Models, as a means to generate images of a class ofobjects and to analyze them, have become increasingly popular. Theproblematic part of this frameworkis the registration of the modelto an image, a.k.a. the fitting. The characteristic features of afitting algorithm are its efficiency, robustness, accuracy andautomation. Many accurate algorithms based on gradient descenttechniques exist which are unfortunately short on the otherfeatures. Recently, an efficient algorithm called InverseCompositional Image Alignment (ICIA) algorithm, able to fit 2Dimages, was introduced. In this paper, we extent this algorithm tofit 3D Morphable Models using a novel mathematical notation whichfacilitates the formulation of the fitting problem. Thisformulation enables us to avoid a simplification so far used in theICIA, being as efficient and leading to improved fitting precision.Additionally, the algorithm is robust without sacrificing itsefficiency and accuracy, thereby conforming to three of the fourcharacteristics of a good fitting algorithm.