Classification Accuracy Improvement for Neuromorphic Computing Systems with One-level Precision Synapses

TL;DR

This paper introduces three methods to enable neuromorphic systems with one-level precision synapses, significantly improving classification accuracy while maintaining high energy efficiency.

Contribution

It proposes distribution-aware quantization, quantization regularization, and bias tuning to enhance synaptic weight resolution in neuromorphic computing.

Findings

Accuracy drop within 0.19% for MNIST

Accuracy drop within 5.53% for CIFAR-10

Methods achieve near-ideal classification performance

Abstract

Brain inspired neuromorphic computing has demonstrated remarkable advantages over traditional von Neumann architecture for its high energy efficiency and parallel data processing. However, the limited resolution of synaptic weights degrades system accuracy and thus impedes the use of neuromorphic systems. In this work, we propose three orthogonal methods to learn synapses with one-level precision, namely, distribution-aware quantization, quantization regularization and bias tuning, to make image classification accuracy comparable to the state-of-the-art. Experiments on both multi-layer perception and convolutional neural networks show that the accuracy drop can be well controlled within 0.19% (5.53%) for MNIST (CIFAR-10) database, compared to an ideal system without quantization.