Deep Spiking Convolutional Neural Network Trained with Unsupervised Spike Timing Dependent Plasticity
本文提出了一个叫 SpiCNN 的神经网络,通过 Leaky-Integrate-and-Fire (LIF) neurons 作为类似激活函数的单元构造神经网络,通过 Spike Timing Dependent Plasticity (STDP) 进行权重学习。
SNN FUNDAMENTALS
Computational Model of Spiking Neuron and Synapse
使用 LIF 模型作为 spiking neuronal dynamics
τ m e m d V m e m d t = − V m e m + W ⋅ δ ( t − t i ) \tau_{mem}\frac{dV_{mem}}{dt}=-V_{mem}+W\cdot \delta(t-t_i) τmemdtdVmem=−Vmem+W⋅δ(t−ti)</