What is a Spiking Neural Network ?

With the increasing demand on big data processing and rising AI technologies, the need for new neural network architectures growed. Regarding this, networks like spiking neural networks (SNNs), which are inspired by the biological behaviour of the brain, are showing interesting features and efficiency in processing. However, what are these and how can they be useful for AI?

As you all know the brain’s capacity and power of learning some things, such as image and voice recognition, is amazing and this process is not working as effectively with computers. In order to solve this issue we should ask what we can learn from our brain structure. 

A brain consists of neurons, which are connected together with synapses. Each neuron can fire an electrical signal through chemical changes in the neuron. At the end with the pattern or frequency of these spikes the learning process starts.

Exactly at this point the idea of spiking neural networks plays a huge role. The idea is implementing a model which realizes the learning process through spikes. In order to that, some approaches are existing, such as latency and rate coding. Latency coding is implemented with the spike timing and rate coding with spike frequency. With these features specific features of the input data can be carried through the characteristics of spike firing.

I hope you enjoyed this article. If you are interested to learn more about it and see more posts to this topic don’t hesitate to contact us.