A Tiny Supervised ODL Core with Auto Data Pruning for Human Activity Recognition

TL;DR

This paper presents a low-power, tiny supervised on-device learning core with automatic data pruning for human activity recognition, reducing power consumption and communication volume with minimal accuracy loss.

Contribution

It introduces an automatic data pruning method integrated with supervised ODL, eliminating manual threshold tuning and enabling efficient on-device learning.

Findings

Power consumption is only 3.39mW.

Data pruning reduces communication volume by 55.7%.

Accuracy loss is only 0.9%.

Abstract

In this paper, we introduce a low-cost and low-power tiny supervised on-device learning (ODL) core that can address the distributional shift of input data for human activity recognition. Although ODL for resource-limited edge devices has been studied recently, how exactly to provide the training labels to these devices at runtime remains an open-issue. To address this problem, we propose to combine an automatic data pruning with supervised ODL to reduce the number queries needed to acquire predicted labels from a nearby teacher device and thus save power consumption during model retraining. The data pruning threshold is automatically tuned, eliminating a manual threshold tuning. As a tinyML solution at a few mW for the human activity recognition, we design a supervised ODL core that supports our automatic data pruning using a 45nm CMOS process technology. We show that the required memory size for the core is smaller than the same-shaped multilayer perceptron (MLP) and the power consumption is only 3.39mW. Experiments using a human activity recognition dataset show that the proposed automatic data pruning reduces the communication volume by 55.7% and power consumption accordingly with only 0.9% accuracy loss.