Neural Living Rooms

Light is sensed by photoreceptors, and changing sound pressure levels induce fluid movements in the cochlea, which are sensed by the deflections of the stereocilia of inner hair cells. These informations are transformed into electrical action potentials and delivered by the optic and acoustic nerves in parallel to different brain areas. In this way, stimuli like moving bars and tones evoke the sensations of motion and pitch. Feature neurons are selectively evoked by spatio-temporally grouped collections of stimuli in their receptive field. The feature attributes span an n-dimensional topological space of the internal object representation. A single spike of a neuron represents one specific feature given by its parameter coordinates and its time of occurrence. Mathematical transforms mediate the mapping from spatio-temporal stimuli to parameterized features. Identified objects are described in instantiations of attribute categories like color, pitch and motion. These color, motion and pitch spaces span a scale from low to high intensity, pitch and speed. Each category then represents a neural living room. Each room is topologically structured with elementary building blocks among repetitive cortical columns and microcircuits. The actions in neural living rooms are orchestrated signaling cascades. For example, a trajectory of spiking neurons represents motion in phase space notation.

The central questions posed for this Research Topic are:
 - How do repetitive stimuli sets group into features?
 - How are these stimuli sets mapped onto internal feature representations?
 - How are these feature maps topologically structured by voting of the stimuli sets?
 - What are active voting, mapping, grouping, pooling, and inhibiting, in this context?
 - Are local trajectories the constituting micro-assets of conservative information in receptive fields?
 - How can biological properties such as voting, mapping, grouping, and inhibiting, be used to improve the information processing ability of current deep neural networks?

The field is widely open to contributions that involve spiking neural networks, polychronous groups, joint weight-delay spike-timing-dependent plasticity, and related topics.