KEYWORDS: Atomic force microscopy, Nose, 3D image processing, 3D modeling, 3D scanning, Image registration, Facial recognition systems, Data modeling, Detection and tracking algorithms, Mouth
The accuracy of a three-dimensional (3-D) face recognition
system depends on a correct registration that aligns the facial
surfaces and makes a comparison possible. The best results obtained
so far use a costly one-to-all registration approach, which
requires the registration of each facial surface to all faces in the
gallery. We explore the approach of registering the new facial surface
to an average face model (AFM), which automatically establishes
correspondence to the preregistered gallery faces. We propose
a new algorithm for constructing an AFM and show that it
works better than a recent approach. We inspect thin-plate spline
and iterative closest-point-based registration schemes under
manual or automatic landmark detection prior to registration. Extending
the single-AFM approach, we consider employing categoryspecific
alternative AFMs for registration and evaluate the effect on
subsequent classification. We perform simulations with multiple
AFMs that correspond to different clusters in the face shape space
and compare these with gender- and morphology-based groupings.
We show that the automatic clustering approach separates the
faces into gender and morphology groups, consistent with the other
race effect reported in the psychology literature. Last, we describe
and analyze a regular resampling method, that significantly increases
the accuracy of registration.
Sequential methods for face recognition rely on the analysis of local facial features in a sequential manner,
typically with a raster scan. However, the distribution of discriminative information is not uniform over the facial
surface. For instance, the eyes and the mouth are more informative than the cheek. We propose an extension
to the sequential approach, where we take into account local feature saliency, and replace the raster scan with
a guided scan that mimicks the scanpath of the human eye. The selective attention mechanism that guides the
human eye operates by coarsely detecting salient locations, and directing more resources (the fovea) at interesting
or informative parts. We simulate this idea by employing a computationally cheap saliency scheme, based on
Gabor wavelet filters. Hidden Markov models are used for classification, and the observations, i.e. features
obtained with the simulation of the scanpath, are modeled with Gaussian distributions at each state of the
model. We show that by visiting important locations first, our method is able to reach high accuracy with much
shorter feature sequences. We compare several features in observation sequences, among which DCT coefficients
result in the highest accuracy.
3D has become an important modality for face biometrics. The accuracy of a 3D face recognition system depends
on a correct registration that aligns the facial surfaces and makes a comparison possible. The best results obtained
so far use a one-to-all registration approach, which means each new facial surface is registered to all faces in
the gallery, at a great computational cost. We explore the approach of registering the new facial surface to an
average face model (AFM), which automatically establishes correspondence to the pre-registered gallery faces.
Going one step further, we propose that using a couple of well-selected AFMs can trade-off computation time
with accuracy. Drawing on cognitive justifications, we propose to employ category-specific alternative average
face models for registration, which is shown to increase the accuracy of the subsequent recognition. We inspect
thin-plate spline (TPS) and iterative closest point (ICP) based registration schemes under realistic assumptions
on manual or automatic landmark detection prior to registration. We evaluate several approaches for the coarse
initialization of ICP. We propose a new algorithm for constructing an AFM, and show that it works better than
a recent approach. Finally, we perform simulations with multiple AFMs that correspond to different clusters in
the face shape space and compare these with gender and morphology based groupings. We report our results on
the FRGC 3D face database.
We propose and compare three different automatic landmarking methods for near-frontal faces. The face information is
provided as 480x640 gray-level images in addition to the corresponding 3D scene depth information. All three methods
follow a coarse-to-fine suite and use the 3D information in an assist role. The first method employs a combination of
principal component analysis (PCA) and independent component analysis (ICA) features to analyze the Gabor feature
set. The second method uses a subset of DCT coefficients for template-based matching. These two methods employ
SVM classifiers with polynomial kernel functions. The third method uses a mixture of factor analyzers to learn Gabor
filter outputs. We contrast the localization performance separately with 2D texture and 3D depth information. Although
the 3D depth information per se does not perform as well as texture images in landmark localization, the 3D information
has still a beneficial role in eliminating the background and the false alarms.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.