Online Supervised Learning for Hardware-Based Multilayer Spiking Neural Networks Through the Modulation of Weight-Dependent Spike-Timing-Dependent Plasticity

Abstract

In this paper, we propose an online learning algorithm for supervised learning in multilayer spiking neural networks (SNNs). It is found that the spike timings of neurons in an SNN can be exploited to estimate the gradients that are associated with each synapse. With the proposed method of estimating gradients, learning similar to the stochastic gradient descent process employed in a conventional artificial neural network (ANN) can be achieved. In addition to the conventional layer-by-layer backpropagation, a one-pass direct backpropagation is possible using the proposed learning algorithm. Two neural networks, with one and two hidden layers, are employed as examples to demonstrate the effectiveness of the proposed learning algorithms. Several techniques for more effective learning are discussed, including utilizing a random refractory period to avoid saturation of spikes, employing a quantization noise injection technique and pseudorandom initial conditions to decorrelate spike timings, in addition to leveraging the progressive precision in an SNN to reduce the inference latency and energy. Extensive parametric simulations are conducted to examine the aforementioned techniques. The learning algorithm is developed with the considerations of ease of hardware implementation and relative compatibility with the classic ANN-based learning. Therefore, the proposed algorithm not only enjoys the high energy efficiency and good scalability of an SNN in its specialized hardware but also benefits from the well-developed theory and techniques of conventional ANN-based learning. The Modified National Institute of Standards and Technology database benchmark test is conducted to verify the newly proposed learning algorithm. Classification correct rates of 97.2% and 97.8% are achieved for the one-hidden-layer and two-hidden-layer neural networks, respectively. Moreover, a brief discussion of the hardware implementations is presented for two mainstream architectures.

A Configurable Event-Driven Convolutional Node with Rate Saturation Mechanism for Modular ConvNet Systems Implementation

“…Despite these clear advantages in neuromorphic systems, conventional frame-based ConvNets are still giving better performance, mostly due to the availability of image datasets mentioned before and very well-known training techniques based on frames, like backpropagation. Nevertheless, recent works have demonstrated similar performance in neuromorphic SNNs using event-based training techniques (Wu et al, 2017 ; Zheng and Mazumder, 2017 ), suggesting that it is only a matter of time that event-based ConvNets become competitive with respect to frame-based ones in terms of classification, while presenting better results in terms of speed and power consumption. Some frame-based approaches are using hardware accelerators (Aydonat et al, 2017 ; Qiao et al, 2017 ) to improve their performance in terms of speed.…”

Neuromorphic LIF Row-by-Row Multi-convolution Processor for FPGA

“…Despite the advantages of neuromorphic systems, accuracies of frame-based accelerators are still better, due to the simple training algorithms, such as backpropagation and the large number of datasets. However, there are recent works in event-based pattern recognition using SCNNs [43] or other techniques, such as HATS [44], which suggest that it is a matter of time until neuromorphic systems become competitive in terms of speed, classification and power consumption.…”

MorphIC: A 65-nm 738k-Synapse/mm$^2$ Quad-Core Binary-Weight Digital Neuromorphic Processor With Stochastic Spike-Driven Online Learning

“…Silicon implementations would come as a last step. In order to overcome the challenges of bottom-up approaches, the development of new multi-layer spike-based learning rules following top-down approaches has gained growing interest in the recent years (e.g., [9]- [12]). Further research is yet required to realize efficient silicon implementations of such learning rules and to make them both compatible with an online-learning setup and able to leverage weight quantization down to binary or ternary resolutions.…”

“…1(b). Spike-based online learning is an active research area, both in the development of new rules for high-accuracy learning in multi-layer networks (e.g., [9]- [12]) and in the demonstration of silicon implementations in applications such arXiv:1904.08513v2 [cs.NE] 16 Jul 2019 as unsupervised learning for image denoising and reconstruction [13], [14]. However, these approaches currently rely on multi-bit weights.…”