In-Memory Resistive RAM Implementation of Binarized Neural Networks for Medical Applications

TL;DR

This paper demonstrates a resistive RAM-based implementation of Binarized Neural Networks tailored for medical applications, focusing on reducing memory and energy use for edge healthcare devices.

Contribution

It introduces a hybrid CMOS-hafnium oxide resistive memory implementation of BNNs and explores strategies to apply them to biomedical signals while maintaining accuracy.

Findings

Memory-accuracy trade-off identified for binarized networks

Effective binarization of classifier part reduces memory without losing accuracy

Results suggest feasibility for edge healthcare devices

Abstract

The advent of deep learning has considerably accelerated machine learning development. The deployment of deep neural networks at the edge is however limited by their high memory and energy consumption requirements. With new memory technology available, emerging Binarized Neural Networks (BNNs) are promising to reduce the energy impact of the forthcoming machine learning hardware generation, enabling machine learning on the edge devices and avoiding data transfer over the network. In this work, after presenting our implementation employing a hybrid CMOS - hafnium oxide resistive memory technology, we suggest strategies to apply BNNs to biomedical signals such as electrocardiography and electroencephalography, keeping accuracy level and reducing memory requirements. We investigate the memory-accuracy trade-off when binarizing whole network and binarizing solely the classifier part. We also discuss how these results translate to the edge-oriented Mobilenet~V1 neural network on the Imagenet task. The final goal of this research is to enable smart autonomous healthcare devices.