Virtually every aspect of our everyday routine actions is embedded in a
sequential context. In this talk I will present and analyze a dynamic
neural field model of ordinal and timing properties of action sequences.
The model extends previous mathematical results on the existence of
multiple bumps of equal strength in dynamic fields to implement a
working memory of sequential events in which varying levels of
self-sustained peak activity are correlated with the relative position
of each item in the sequence. Our ultimate goal is to validate the
sequence model as part of an existing dynamic field architecture for
human-robot interaction (HRI). |