Building Robust Systems out of Non-Robust Components
Pradeep K. Khosla
Abstract:
Most of current machine vision systems suffer
from a lack of flexibility to account for the high variability of unstructured
environments. As the state of the world evolves the knowledge provided by
different visual attributes changes, breaking the initial assumptions of the
vision system. This results in the vision algorithm/system not being able to
accomplish its original goals. We have
developed a new approach for the creation of an adaptive visual system that
demonstrates robust behavior even when its components are non-robust. The system is able to selectively combine
information from different visual algorithms depending on its evaluation of
which algorithms are making a contribution to its goal. Using a probabilistic approach and
uncertainty metrics, the system is able to take appropriate decisions about the
more relevant visual attributes to consider. The system is based on an
intelligent agent paradigm. Each visual algorithm is implemented as an agent,
which adapts its behavior according to uncertainty considerations. The proposed
system aims to achieve robustness and efficiency. By combining the outputs of
multiple vision modules the assumptions and constraints of each module are
factored out resulting in a more robust system. Efficiency is achieved through
the on-line selection and specialization of the agents. An implementation of the
system for the task of human tracking has demonstrated excellent results.
This is joint work with Alvaro Soto.
Brief
Biography:
Pradeep K. Khosla
is the Philip and Marsha Dowd Professor in the