CORTEX 450 Biologically Plausible Cortical Neural Network with Visual Recognition, Synthetic EEG, Autonomous Navigation and Visual Neuroprosthetics Perspectives

Cortex 450 is a software simulator of a cortical neural network integrating biophysical Leaky Integrate-and-Fire (LIF) neurons with STDP synaptic plasticity, real image recognition via a multi-feature encoder, synthetic EEG signal generation across five brain frequency bands (delta, theta, alpha, beta, gamma), 2D map-based autonomous navigation, and a semantic decoding system based on a hybrid image–neural-state signature.The system learns image–concept associations within a few dozen simulation steps thanks to the hybrid signature (85% image vector + 15% neural state) and operates in real time at a 14× RT ratio on standard hardware. The vectorised LIF engine with sparse synaptic matrix achieves 8,000 steps/second in navigation mode.This paper describes the architecture, experimental results, and multiple future research directions, including optic nerve stimulation via wave patterns generated by the simulated cortical network — an original hypothesis that opens a new research space in the field of visual neuroprosthetics