Title: Grossberg Models
Dr Noel Murphy
Short Summary: Interactive
workshop on developing the application of Grossberg models of visual processing
to practical image and video analysis problems.
Long Summary: Over
many years the work of Stephen Grossberg and his colleagues on models of
biological vision has given a unique insight to the computational processes
that underlie this vital sense. Despite the enormous leap in understanding
contained therein, the distributed, continuous time PDE-based models are
still very far from practical implementation in real machine vision and
computer vision applications. The aim of this workshop is to develop the
participant's understanding of the concepts involved in Grossberg's models
and to begin the process of bridging the gap between model and application.
The concepts involved include those of the boundary contour system and
feature contour system for visual form analysis (segmentation), the computational
roles of resonance and reset in these models and the different computation
rules which require parallel systems for processing static and moving form.
It is intended that the workshop would involve an interactive and exploratory
development of ideas, rather than be purely expositional.