TR2005-093

A Bilinear Illumination Model for Robust Face Recognition


    •  Jinho Lee, Baback Moghaddam, Hanspeter Pfister, Raghu Machiraju, "A Bilinear Illumination Model for Robust Face Recognition", Tech. Rep. TR2005-093, Mitsubishi Electric Research Laboratories, Cambridge, MA, August 2005.
      BibTeX TR2005-093 PDF
      • @techreport{MERL_TR2005-093,
      • author = {Jinho Lee, Baback Moghaddam, Hanspeter Pfister, Raghu Machiraju},
      • title = {A Bilinear Illumination Model for Robust Face Recognition},
      • institution = {MERL - Mitsubishi Electric Research Laboratories},
      • address = {Cambridge, MA 02139},
      • number = {TR2005-093},
      • month = aug,
      • year = 2005,
      • url = {https://www.merl.com/publications/TR2005-093/}
      • }
  • Research Area:

    Computer Vision

Abstract:

We present a technique to generate an illumination subspace for arbitrary 3D faces based on the statistics of measured illuminations under variable lighting conditions from many subjects. A bilinear model based on the higher-order singular value decomposition is used to create a compact illumination subspace given arbitrary shape parameters from a parametric 3D face model. Using a fitting procedure based on minimizing the distance of the input image to the dynamically changing illumination subspace, we reconstruct a shape-specific illumination subspace from a single photograph. We use the reconstructed illumination subspace in several face recognition experiments with variable lighting conditions and obtain accuracies that are very competitive with previous methods that require training sessions or multiple images of the subject.