Biological motivation for neural networks

Neuron activity in the human brain is typically excited or inhibited through connections to other neurons. The fastest neuron switching times are known to be of order $10^{-3}$ seconds [7, 4.1.1]. This is quite slow compared to about $10^{-10}$ for a computer. Yet, we are able to recognize something like our mother in less than $10^{-1}$ seconds. Thus, this process cannot take more than about $10^2 = 100$ steps knowing the switching time of a single neuron. This suggests that our neural system must follow highly parallel process distributed over many neurons. This biological evidence motivates the artificial neural network model in visual perception.