Deep Transfer Learning with Graph Neural Network for Sensor-Based Human Activity Recognition

TL;DR

This paper introduces a graph neural network-based transfer learning approach for sensor-based human activity recognition, effectively addressing data scarcity and sensor variation issues with high accuracy and good transferability.

Contribution

The paper proposes a novel ResGCNN model with a multi-layer residual structure for sensor-based HAR, demonstrating superior transferability and few-shot learning capabilities.

Findings

Achieved over 98% accuracy on PAMAP2 and mHealth datasets.

ResGCNN outperforms existing models in transfer learning scenarios.

Shows strong meta-learning ability for sensor-based HAR.

Abstract

The sensor-based human activity recognition (HAR) in mobile application scenarios is often confronted with sensor modalities variation and annotated data deficiency. Given this observation, we devised a graph-inspired deep learning approach toward the sensor-based HAR tasks, which was further used to build a deep transfer learning model toward giving a tentative solution for these two challenging problems. Specifically, we present a multi-layer residual structure involved graph convolutional neural network (ResGCNN) toward the sensor-based HAR tasks, namely the HAR-ResGCNN approach. Experimental results on the PAMAP2 and mHealth data sets demonstrate that our ResGCNN is effective at capturing the characteristics of actions with comparable results compared to other sensor-based HAR models (with an average accuracy of 98.18% and 99.07%, respectively). More importantly, the deep transfer learning experiments using the ResGCNN model show excellent transferability and few-shot learning performance. The graph-based framework shows good meta-learning ability and is supposed to be a promising solution in sensor-based HAR tasks.