Resource and Energy Efficient Implementation of ECG Classifier using Binarized CNN for Edge AI Devices

Wearable Artificial Intelligence-of-Things (AIoT) devices demand smart gadgets that are both resource and energy-efficient. In this paper, we explore efficient implementation of binary convolutional neural network employing function merging and block reuse techniques. The hardware implemented in field programmable gate array (FPGA) platform can classify ventricular beat in electrocardiogram achieving accuracy of 97.5%, sensitivity of 85.7%, specificity of 99.0%, precision of 92.3%, and F1-score of 88.9% while consuming only 10.5-mu W of dynamic power dissipation.

Automated summary

This paper explores an efficient implementation of binary convolutional neural network using function merging and block reuse techniques, achieving accurate classification of ventricular beat in electrocardiogram with low energy consumption on FPGA platform.