Marni Stewart Bartlett
Marian Stewart Bartlett, Ph.D.
Marian Stewart Bartlett
The Salk Institute, CNL
10010 North Torrey Pines Road
La Jolla, CA 92037
619-453-4100, ext. 1420
I have moved to the University of California, San Diego, where I am
doing a postdoc in the Machine Perception Lab with Javier Movellan. New web page and contact information.
I am a postdoc with Terry Sejnowski at the Salk Institute working on
image analysis and statistical pattern recognition. I received
my PhD in Cognitive Science and Psychology from the University of
California, San Diego in 1998. My doctoral dissertation was on "Face image
analysis by unsupervised learning and redundancy reduction."
1.9 megs compressed.
View a full list of my
My papers on
ICA for face image analysis.
Download my academic CV.
CNL home page.
Other papers from my lab can be found on
CNL publications page or
Links to UCSD Departments: Psychology,
View web statistics.
1. Facial expression recognition
The Facial Action Coding System, (FACS), devised by Ekman and Friesen,
provides an objective means for measuring the facial muscle contractions
involved in a facial expression. In this paper, we approach automated
facial expression analysis by detecting and classifying facial actions. We
generated a database of over 1100 image sequences of 24 subjects performing
over 150 distinct facial actions or action combinations. We compare three
different approaches to classifying the facial actions in these images:
Holistic spatial analysis based on principal components of graylevel
images; explicit measurement of local image features such as wrinkles; and
template matching with motion flow fields. On a dataset containing six
individual actions and 20 subjects, these methods had 89%, 57%, and 85%
performances respectively for generalization to novel subjects. When
combined, performance improved to 92%.
Bartlett, M.S., Hager, J.C., Ekman, P., and Sejnowski,
T.J. (1999). Measuring facial expressions by computer image
analysis. Psychophysiology, 36, p. 253-263.
Bartlett, M.S., Viola, P. A., Sejnowski, T. J.,
Golomb, B.A., Larsen, J., Hager, J. C., and Ekman, P. (1996). Classifying facial action, Advances in Neural Information
Processing Systems 8, MIT Press, Cambridge, MA. p. 823-829.
Our more recent work explores and compares over 7 different techniques for
facial expression analysis, including analysis of facial motion
through estimation of optical flow; holistic spatial analysis such as
principal component analysis, independent component analysis, local
feature analysis, and linear discriminant analysis; and methods based on
the outputs of local filters, such as Gabor wavelet representations, and
local principal components. Performance of these systems is compared to
naive and expert human subjects. Best performances were obtained using
the Gabor wavelet representation and the independent component
representation, both of which achieved 96% accuracy for classifying
twelve facial actions of the upper and lower face. The results provide
converging evidence for the importance of using local filters, high
spatial frequencies, and statistical independence for classifying facial
Donato, G.L., Bartlett, M.S., Hager, J.C., Ekman, P., and
Sejnowski, T.J. (1999). Classifying Facial Actions IEEE
Transactions on Pattern Analysis and Machine Intelligence 21(10) p. 974-989.
2. Independent component analysis for representing and
In a task such as face recognition, much of the important information
may be contained in the high-order relationships among the image pixels.
Some success has been attained using data-driven face representations based
on principal component analysis, such as "Eigenfaces" (Turk & Pentland,
1991) and "Holons" (Cottrell & Metcalfe, 1991). Principal component
analysis (PCA) is based on the second-order statistics of the image set,
and does not address high-order statistical dependencies such as the
relationships among three or more pixels. Independent component analysis
(ICA) is a generalization of PCA which separates the high-order moments of
the input in addition to the second-order moments.
We developed image representations based on the independent
components of the face images and compared them to a PCA representation for
face recognition. ICA was performed on a set of face images by an
unsupervised learning algorithm derived from the principle of optimal
information transfer through sigmoidal neurons (Bell & Sejnowski,
1995). The algorithm maximizes the mutual information between the input and
the output, which produces statistically independent outputs under certain
ICA was performed on the face images under two different
architectures. The first architecture provided a set of statistically
independent basis images for the faces that can be viewed as a set of
independent facial features. These ICA basis images were spatially
local, unlike the PCA basis vectors. The representation consisted of the
coefficients for the linear combination of basis images that comprised each
face image. The second architecture produced independent coding variables
(coefficients). This provided a factorial face code, in which the
probability of any combination of features can be obtained from the product
of their individual probabilities. The distributions of these coefficents
were sparse and highly kurtotic. Classification was performed using nearest
neighbor, with similarity measured as the cosine of the angle between
representation vectors. Both ICA representations were superior to the PCA
representation for recognizing faces across sessions, changes in
expression, and changes in pose.
M. Stewart, Lades, H. Martin, and Sejnowski, T.J. (1998). Independent
component representations for face recognition. Proceedings of the
SPIE, Vol 3299: Conference on Human Vision and Electronic Imaging III,
p. 528-539. Download
Bartlett, M. Stewart, and Sejnowski,
T. J. (1997). Independent components of face images: A representation for
face recognition. Proceedings of the 4th Annual Jount Symposium on Neural
Computation, Pasadena, CA, May 17, 1997. Proceedings can be obtained from
the Institute for Neural Computation, UCSD 0523, La Jolla, CA 92093. Download
Bartlett, M. Stewart, and Sejnowski, T. J. (1997). Viewpoint
invariant face recognition using independent component analysis and
attractor networks. In M. Mozer, M. Jordan, T. Petsche (Eds.), Advances
in Neural Information Processing Systems 9, MIT Press, Cambridge,
MA. 817-823. Download
3. Unsupervised learning of invariant representations of faces
The appearance of an object or a face changes continuously as the observer
moves through the environment or as a face changes expression or pose.
Recognizing an object or a face despite these image changes is a
challenging problem for computer vision systems, yet we perform the task
quickly and easily. This simulation investigates the ability of an
unsupervised learning mechanism to acquire representations that are
tolerant to such changes in the image. The learning mechanism finds these
representations by capturing temporal relationships between 2-D patterns.
Previous models of temporal association learning have used idealized input
representations. The input to this model consists of graylevel images of
faces. A two-layer network learned face representations that incorporated
changes of pose up to 30 degrees. A second network learned
representations that were independent of facial expression (Bartlett &
In the next phase of this work, we investigated the ability of an attractor
network to acquire view invariant visual representations by associating
first neighbors in a pattern sequence. The pattern sequence contained
successive views of faces of ten individuals as they change pose. Under
the network dynamics developed by Griniasty, Tsodyks \& Amit (1993),
multiple views of a given subject fell into the same basin of attraction,
producing viewpoint invariant representations (Bartlett & Sejnowski,
We next derived a generalization of the attractor network learning rule in
Griniasty et al. (1993). We showed that combining temporal smoothing of
unit activities with basic Hebbian learning in a Hopfield network produces
a generalization of the learning rule in Griniasty et al. that associates
temporally proximal inputs rather than just first neighbors in the pattern
sequence. This learning rule achieved greater viewpoint invariance than
that of Griniasty et al. (Bartlett \& Sejnowski, 1998; 1997).
Bartlett, M.S., and Sejnowski, T.J. (1998). Learning
viewpoint invariant face representations from visual experience in an attractor
network. Network: Computation in Neural Systems 9(3) 399-417.
Bartlett, M. Stewart, and Sejnowski, T.J. (1998). Learning Viewpoint Invariant Face Representations from Visual
Experience by Temporal Association. In H. Wechsler, P.J. Phillips,
V. Bruce, S. Fogelman-Soulie, T. Huang (Eds.), Face Recognition: From
Theory to Applications, NATO ASI Series F. Springer-Verlag. Download
Bartlett, M. Stewart, and Sejnowski, T. J. (1997). Viewpoint invariant face recognition using independent component
analysis and attractor networks. In M. Mozer, M. Jordan, T. Petsche
(Eds.), Advances in Neural Information Processing Systems 9, MIT
Press, Cambridge, MA. 817-823. Download
Bartlett, M. Stewart, and Sejnowski, T. J. (1996b).
Learning Viewpoint Invariant Representations of Faces in an Attractor
Network. Poster presented at the 18th Cognitive Science Society Meeting,
San Diego, CA, July 12-15, 1996.
Bartlett, M. Stewart, and Sejnowski, T. J. (1996a). Unsupervised learning of invariant representations of faces through
temporal association. In Computational Neuroscience: International
Review of Neurobiology Suppl. 1. J.M. Bower, ed. Academic Press, San Diego,
CA., 1996. p. 317-322.
Number of hits since July 10, 1998: