Acquiring Linear Subspaces for Face Recognition under
Variable Lighting
Abstract
Previous work has demonstrated that the image variation of many
objects (human faces in particular) under variable lighting can be
effectively modeled by low dimensional linear spaces, even when
there are multiple light sources and shadowing. Basis images
spanning this space are usually obtained in one of three ways: A
large set of images of the object under different lighting
conditions is acquired, and principal component analysis (PCA) is
used to estimate a subspace. Alternatively, synthetic images are
rendered from a 3D model (perhaps reconstructed from images) under
point sources, and again PCA is used to estimate a subspace.
Finally, images rendered from a 3D model under diffuse lighting
based on spherical harmonics are directly used as basis images. In
this paper, we show how to arrange physical lighting so that the
acquired images of each object can be directly used as the basis
vectors of a low-dimensional linear space, and that this subspace
is close to those acquired by the other methods. More
specifically, there exist configurations of k point light source
directions, with k typically ranging from 5 to 9, such that by
taking k images of an object under these single sources, the
resulting subspace is an effective representation for recognition
under a wide range of lighting conditions. Since the subspace is
generated directly from real images, potentially complex and/or
brittle intermediate steps such as 3D reconstruction can be
completely avoided; nor is it necessary to acquire large numbers
of training images or to physically construct complex diffuse
(harmonic) light fields. We validate the use of subspaces
constructed in this fashion within the context of face
recognition.
Results [Finding Universal Configuration]
Left: The universal configuration of nine light source
directions with all 200 sample points plots on a hemisphere.
Right: Nine images of a person illuminated by lights from the universal configuration.
[Face Recognition by Universal Configuration, 9PL]
The error rates for various recognition methods on subsets of the Yale Face Database B. Some of the
entries (indicated by citation) were taken from published papers
whereas the 9PL, Harmonic Images, and Nearest Neighbor results are
from our own implementation.
Reference
Acquiring Linear Subspaces for Face Recognition under Variable Lighting [ pdf ]
Kuang-Chih Lee, Jeffrey Ho, David Kriegman / submitted to IEEE Trans. Pattern Analysis and Machine Intelligence, 2003, second review
Nine Points of Lights: Acquiring Subspaces for Face Recognition under Variable Illumination [ pdf ]
Kuang-Chih Lee, Jeffrey Ho, David Kriegman / IEEE Conf. On Computer Vision and Pattern Recognition, 2001, oral presentation, vol. 1, pp. 519-526
On Reducing the Complexity of Illumination Cones [ pdf ]
Jeffrey Ho, Kuang-Chih Lee, David Kriegman / IEEE Workshop on Identifying Objects Across Variations in Lighting: Psychophysics and Computation, 2001, pp. 56-63
