Multi-channel Time Series Decomposition Network For Generalizable Sensor-Based Activity Recognition

TL;DR

This paper introduces MTSDNet, a novel sensor-based activity recognition model that decomposes signals into components for improved cross-person generalization, achieving higher accuracy and interpretability across multiple datasets.

Contribution

The paper proposes a new multi-channel time series decomposition network that enhances out-of-domain generalization in activity recognition by learning low-rank representations through trainable decomposition.

Findings

Outperforms existing methods in accuracy on multiple datasets

Demonstrates improved stability and interpretability

Effective in cross-person and out-of-domain scenarios

Abstract

Sensor-based human activity recognition is important in daily scenarios such as smart healthcare and homes due to its non-intrusive privacy and low cost advantages, but the problem of out-of-domain generalization caused by differences in focusing individuals and operating environments can lead to significant accuracy degradation on cross-person behavior recognition due to the inconsistent distributions of training and test data. To address the above problems, this paper proposes a new method, Multi-channel Time Series Decomposition Network (MTSDNet). Firstly, MTSDNet decomposes the original signal into a combination of multiple polynomials and trigonometric functions by the trainable parameterized temporal decomposition to learn the low-rank representation of the original signal for improving the extraterritorial generalization ability of the model. Then, the different components obtained by the decomposition are classified layer by layer and the layer attention is used to aggregate components to obtain the final classification result. Extensive evaluation on DSADS, OPPORTUNITY, PAMAP2, UCIHAR and UniMib public datasets shows the advantages in predicting accuracy and stability of our method compared with other competing strategies, including the state-of-the-art ones. And the visualization is conducted to reveal MTSDNet's interpretability and layer-by-layer characteristics.