A wave-pulse neural network for quasi-quantum coding
We design a physical wave-pulse neural network (WPNN) for both wave and pulse propagation, which gives more degrees of freedom for neural coding than spike neural networks (SNN). We define the rules and the information entropy of this kind of neural network, where the signal speed, arrival time, and the length of connections between neurons all become crucial parameters for signal coding. We call it quasi-quantum coding (QQC) since the combination of wave and pulse signals here behaves like a classical mimic of quantum wave-particle duality, and can be studied by borrowing some concepts form quantum mechanics. We present that the quasi-quantum coding can give efficient methods for both sound and image recognitions. We also discuss the possibility of the wave-pulse neural network and the quasi-quantum coding methods running on it in biological brains where both neural oscillations and action potentials are important to cognition.