Computation through the development of synchrony in neural models

We investigate integrate-and-fire models with an emphases on temporal order. Our goal is to demonstrate that simple models can perform non- trivial computations based on the development of synchrony. While the computational strategies are general, we realize our model in a sensory processing context, particularly olfaction, where we consider recognizing an odor in the presence of a background, using multiple sniffs. This introduces a time-dependent aspect in the stimuli where the advantages of temporal coding may become more apparent. Our synchrony-based computational strategy emphasizes the role of adaptation and also employs the general principle: ``What moves together is an object.''