Keynote Speakers
- Alex Pentland, MIT Media Lab, USA
- Michael J. Black, Xerox Palo Alto Research Center, USA
Probabilistic Methods for Representing, Tracking and Understanding
Human
Motion
Abstract
We present a method for the modeling and tracking of human motion
in a sequence of 2D video images. Our analysis is divided in two
parts: First, we estimate a statistical model of typical activities
from a large set of 3D human motion data. In a second step we use this
probabilistic model as a prior distribution for Bayesian propagation
using particle filtering.
From a statistical modeling perspective, a 3D human motion can be
thought of as a collection of time-series. The human body is
represented as a set of articulated cylinders with 25 degrees of
freedom and the evolution of a particular joint angle is described by
one of the time-series. A key difficulty for the modeling of these
data is that each time-series has to be decomposed into suitable
temporal primitives prior to statistical analysis. We have been
developing methods for the automatic segmentation and modeling of such
motion primitives using techniques from statistical learning theory.
Learned temporal models provide prior probability distributions
over human motions which can be used in a Bayesian framework for
tracking. For this purpose, we specify a generative model of image
appearance and the likelihood of observing image data given the model.
The high dimensionality and non-linearity of the articulated human
body model and the ambiguities in matching the generative model to the
image result in a posterior distribution that cannot be represented in
closed form. Hence, the posterior is represented using a discrete set
of samples and is propagated over time using particle filtering. The
learned temporal prior helps constrain the sampled posterior to
regions of the parameter space with a high probability of
corresponding to human motions. The resulting algorithm is able to
track human subjects in monocular video sequences and recover their 3D
motion under changes in their pose and against complex unknown
backgrounds.
Joint work with:
Hedvig Sidenbladh (Royal Institute of Technology, Sweden)
David Fleet (Xerox PARC)
Dirk Ormoneit (Stanford, Dept of Statistics)
- Horst Bunke, Department of Computer Science, University of Bern,
Switzerland
Graph Matching - Recent Theoretical Results, Algorithms, and
Applications
Abstract
Graphs are a powerful and versatile tool useful in various subfields of
science
and engineering. In many applications, for example, in pattern
recognition and computer vision, it is required to measure the similarity
of objects. When graphs are used for the representation of structured
objects, then the problem of measuring object similarity turns into the
problem of computing the similarity of graphs, which is also known as
graph matching. In this session, similarity measures on graphs and related
algorithms will be reviewed. Applications of graph matching will be
demonstrated by several examples from the fields of pattern recognition and
computer vision. Also, recent theoretical work showing various
relations between different similarity measures will be discussed.