Multidimensional Morphable Models: A Framework for Representing and Matching Object Classes

  • Authors:
  • Michael J. Jones;Tomaso Poggio

  • Affiliations:
  • Cambridge Research Lab, Digital Equipment Corp., One Kendall Sq., Bldg 700, Cambridge, MA 02139. E-mail: mjones@crl.dec.com;Artificial Intelligence Lab and The Center for Biological and Computational Learning, Massachusetts Institute of Technology, Cambridge, MA 02139. E-mail: tp@ai.mit.edu

  • Venue:
  • International Journal of Computer Vision
  • Year:
  • 1998

Quantified Score

Hi-index 0.00

Visualization

Abstract

We describe a flexible model for representing images of objects of acertain class, known a priori, such as faces, and introduce a newalgorithm for matching it to a novel image and thereby perform image analysis. The flexible model, known as a multidimensional morphable model, is learned from example images of objects of aclass. In this paper we introduce an effective stochastic gradientdescent algorithm that automatically matches a model to a novel image.Several experiments demonstrate the robustness and the broad range ofapplicability of morphable models. Our approach can provide novelsolutions to several vision tasks, including the computation of imagecorrespondence, object verification and imagecompression.