Sequence Approximation using Feedforward Spiking Neural Network for Spatiotemporal Learning: Theory and Optimization Methods
A dynamical system of spiking neurons with only feedforward connections can classify spatiotemporal patterns without recurrent connections. However, the theoretical construct of a feedforward Spiking Neural Network (SNN) for approximating a temporal sequence remains unclear, making it challenging to optimize SNN architectures for learning complex spatiotemporal patterns. In this work, we establish a theoretical framework to understand and improve sequence approximation using a feedforward SNN. Our framework shows that a feedforward SNN with one neuron per layer and skip-layer connections can approximate the mapping function between any arbitrary pairs of input and output spike train on a compact domain. Moreover, we prove that heterogeneous neurons with varying dynamics and skip-layer connections improve sequence approximation using feedforward SNN. Consequently, we propose SNN architectures incorporating the preceding constructs that are trained using supervised backpropagation-through-time (BPTT) and unsupervised spiking-timing-dependent plasticity (STDP) algorithms for classification of spatiotemporal data. A Dual Search-space Bayseian Optimization method is developed to optimize architecture and parameters of the proposed SNN with heterogeneous neuron dynamics and skip-layer connections.
PDF AbstractDatasets
Task | Dataset | Model | Metric Name | Metric Value | Global Rank | Benchmark |
---|---|---|---|---|---|---|
Gesture Recognition | DVS128 Gesture | mMND (STDP) | Accuracy (%) | 96.6 | # 9 | |
Gesture Recognition | DVS128 Gesture | mMND (BPTT) | Accuracy (%) | 98.0 | # 2 | |
Image Classification | N-Caltech 101 | mMND (STDP) | Accuracy | 58.1 | # 1 | |
Image Classification | N-Caltech 101 | mMND (BPTT) | Accuracy | 71.2 | # 2 |