A Multi-Layer Parallel LSTM Network for Human Activity Recognition with Smartphone Sensors

Abstract

With the development of mobile communication human activity recognition (HAR) with smartphones has attracted a lot of attentions in recent years. On the other hand, the appearance of deep learning technologies makes it possible to extract features automatically instead of hand-crafted extracting features in the traditional machine learning methods. Among deep model, CNN-based HAR methods dominate the studies compared to RNN-based methods. In this paper, we propose a RNN-based multi-layer parallel LSTM network to recognize human activities. The experimental results on the public UCI HAR dataset indicate that the proposed approach performs better than the traditional machine-learning methods, and achieves the similar performance as that of CNN, but it has lower computation complexity than CNN-based methods.