Dynamics of functional coupling in the cerebral cortex: an attempt at a model-based interpretation
Proceedings of the NATO advanced research workshop and EGS topical workshop on Chaotic advection, tracer dynamics and turbulent dispersion
Neural Assemblies, an Alternative Approach to Artificial Intelligence
Neural Assemblies, an Alternative Approach to Artificial Intelligence
Towards Novel Neuroscience-Inspired Computing
Emergent Neural Computational Architectures Based on Neuroscience - Towards Neuroscience-Inspired Computing
| Hi-index | 0.00 |
For the recognition of objects there are a number of computational requirements that go beyond the detection of simple geometric features like oriented lines. When there are several partially occluded objects present in a visual scene one has to have an internal knowledge about the object to be identified, e.g. using associative memories. We have studied the bidirectional dynamical interaction of two areas, where the lower area is modelled to match area V1 in greater detail and the higher area uses Hebbian learning to form an associative memory for a number of geometric shapes. Both areas are modelled with simple spiking neuron models, and questions of "binding" by spike-synchronisation and of the effects of Hebbian learning in various synaptic connections (including the long-range cortico-cortical projections) are studied. Presenting a superposition of three stimulus objects corresponding to learned assemblies, we found generally two states of activity: (i) relatively slow and unordered activity, synchronized only within small regions, and (ii) faster oscillations, synchronized over larger regions. The neuron groups representing one stimulus tended to be simultaneously in either the slow or the fast state. At each particular time, only one assembly was found to be in the fast state. Activation of the three assemblies switched within a few hundred milliseconds.