A Spiking Neuron Model of Head-Direction Cells for Robot Orientation

This paper proposes a bio-mimetic model of headdirection (HD) cells implemented on a real robot. The model is based on spiking neurons to study the temporal aspects of state transitions of the HD cell activity following reorienting visual stimuli. The short transient latencies observed experimentally are reproduced by the model. We focus on the integration of angular velocity inertial signals provided by accelerometers. This integration is realized by a continuous attractor network modeling the interaction between the lateral mammillary nucleus (LMN) and the dorsal tegmental nucleus (DTN), two structures belonging to the HD cell anatomical circuit. Relevant parameters defining the connections between LMN and DTN are determined by a genetic algorithm.