Efficient Spiking Neural Networks with Radix Encoding
This addresses efficiency issues in neuromorphic computing for applications like edge devices, but it is incremental as it builds on existing SNN methods.
The paper tackles the problem of spiking neural networks (SNNs) requiring long spike trains for accuracy, which reduces efficiency, by proposing a radix encoded SNN with ultra-short spike trains. The result is a 25X speedup and a 1.1% accuracy increase compared to state-of-the-art methods on VGG-16 and CIFAR-10.
Spiking neural networks (SNNs) have advantages in latency and energy efficiency over traditional artificial neural networks (ANNs) due to its event-driven computation mechanism and replacement of energy-consuming weight multiplications with additions. However, in order to reach accuracy of its ANN counterpart, it usually requires long spike trains to ensure the accuracy. Traditionally, a spike train needs around one thousand time steps to approach similar accuracy as its ANN counterpart. This offsets the computation efficiency brought by SNNs because longer spike trains mean a larger number of operations and longer latency. In this paper, we propose a radix encoded SNN with ultra-short spike trains. In the new model, the spike train takes less than ten time steps. Experiments show that our method demonstrates 25X speedup and 1.1% increment on accuracy, compared with the state-of-the-art work on VGG-16 network architecture and CIFAR-10 dataset.