|
Color
Texture Feature for Robust Face Recognition |
|
1. Introduction
l
Color information can
provide complementary information to face recognition (FR) using texture
analysis.
l
Most of FR approaches
using color texture has been limited to encoding the texture patterns of only
color pixel variations derived from each individual spectral band image.
l
We propose a new face
descriptor so-called Local Color Vector Binary Patterns (LCVBP) for a FR
purpose.
LCVBP consists of two
discriminant patterns: color norm patterns and color angular patterns.
In particular, color
angular patterns allow for encoding the discriminating texture patterns that
are derived from spatial interactions among different spectral band images.
2. Extraction of Local Color Vector
Binary Patterns
Fig. 1. Defining
the color vector of a pixel location within local regions of three channels in
YIQ color space.
Fig. 2. Illustration
of extracting the LCVBP from YIQ color image.
l
Step 1: Define color vector at a center pixel location of each local region obtained
from a set of color band images
.
l
Step 2: Compute color norm
value and color angle values (between i-th
and j-th bands) of , respectively.
l
Step 3: Define uniform LBP
operators and for a color norm pattern and
color angular patterns at location .
|
If |
(1) |
Otherwise |
: Color norm values of color vector that form a circular
neighborhood
: Indexing function that labels a particular index to each of uniform
patterns
: Total number of transitions at from 0 to 1 or vice versa
are defined in a similar way
by using the Equation of
l
Step 4: Compute LBP
histograms of each local region for color norm and color
angular patterns.
l
Step 5: Generate LBP
histograms of for color norm and color
angular patterns by separately concatenating regional LBP histograms obtained
from associated patterns.
3. Face Recognition with LCVBP
Feature
l
Low-dimensional feature
extraction technique is adopted to protect FR performance from being degraded
caused by high dimensionality and redundant information
For |
(2) |
: Feature extractors formed with LBP histograms of training face images
for color norm and color angular patterns
: LBP histograms of a probe RGB face image for color norm pattern and
color angular patterns
: Low-dimensional features of and
l
Feature-level
information fusion technique is used to combine a total of features (one for color norm
and for color angular patterns)
for facilitating complementary effect
|
(3) |
l
LCVBP feature of a
gallery RGB face image is created in the same way as probe RGB face image.
l
A nearest neighbor
(NN) classifier is used for FR.
4. Experiment
<Experimental setup>
l For comparison, FR methods using only grayscale texture (i.e., grayscale
LBP, grayscale Gabor) and FR method using color texture (i.e., color LBP) were
used
l RQCr (R from RGB, and Q and Cr from YIQ and YCbCr
color spaces) color configuration was used for extracting the LCVBP feature.
|
(a) |
|
(b) |
|
(c) |
Fig. 3. Examples of used face
images. (a) CMU-PIE DB (with illumination change). (b) XM2VTS DB (with
illumination change). (C) Color FERET DB (facial pose change).
<Experimental
results>
Fig. 4. Results on
illumination variation (left) and pose variation (right).
l Under severe illumination variation
A total of 2,492 illuminant face images of 201 subjects were collected
from the CMU-PIE and XM2VTS DB.
The LCVBP feature is more robust against illumination variation than the
grayscale LBP, grayscale Gabor, and color LBP.
l Under pose variation
A total of 1,378 face images of 107 subjects were collected from the Color
FERET.
The LCVBP feature is effective for FR under pose variation
5. Conclusions
l In the proposed LCVBP, color norm and color angular patterns are extracted
from a color face image
l
The LCVBP is highly
robust to severe variation in illumination and moderate facial pose changes
* Contact Person: Prof. Yong Man Ro (ymro@kaist.ac.kr)
1. S. H. Lee, J. Y. Choi, Y. M. Ro, and K. N. Plataniotis, ¡°Local Color Vector Binary Patterns from Multichannel Face Images for Face Recognition,¡± IEEE Transactions on Image Processing, 2012.
2. J. Y. Choi, Y. M. Ro, and K. N. Plataniotis, ¡°Color Local Texture Features for Color Face Recognition,¡± IEEE Transactions on Image Processing, 2012.
3. S. H. Lee, H. Kim, K. N. Plataniotis, and Y. M. Ro, ¡°Using Color Texture Sparsity for Facial Expression Recognition,¡± IEEE International Conference on Automatic Face and Gesture Recognition, 2013.
4. S. H. Lee, H. Kim, Y. M. Ro, ¡°A Comparative Study of Color Texture Features for Face Analysis,¡± IAPR The Fourth Computational Color Imaging Workshop, 2013.
5. S. H. Lee, J. Y. Choi, K. N. Plataniotis, and Y. M. Ro, ¡°Local Color Vector Binary Patterns for Face Recognition,¡± IEEE Conference on Image Processing, 2011.
l Video1: Effectiveness of color
texture under illumination change.
l Video2: Effectiveness of color
texture for small resolution faces.