pyBrain

A multi-model-supporting framework implemented in Python for SNN simulating.

Background

A brief introduction to SNN can be seen here: INTRODUCTION (Chinese version)

Supported / Supporting Features

Neuron Models

• Izhikevich neuron
• H-H neuron: low performance so perhaps it would be implemented after a long time XD)
• LIF neuron: fast but low biological plausibility

Synapses / Receptors

• AMPA receptor
• NMDA receptor
• GABA_A receptor
• GABA_B receptor
• electrical synapse

Simulation Methods (Integrators)

• clock-driven method with high precision spiking time searching
• voltage-driven: the best method so far
• event-driven (spiking-driven): best performance but cannot be generally used

Training Algorithms

• unsupervised STDP
• supervised STDP (ReSuMe)

Spike Coding

I am still studying it.

Others

• oscilloscopes
• GPU acceleration