On Training Strategies for LSTMs in Sensor-Based Human Activity Recognition

Human Activity Recognition (HAR) is one of the core research areas in mobile and wearable computing. With the application of deep learning (DL) techniques such as CNN, recognizing periodic/static activities (e.g, walking, lying, cycling, etc.) has become a well-studied problem. What remains a major challenge though is the sporadic activity recognition (SAR) problem, where activities of interest tend to be non-periodic, and occur less frequently when compared with the often large amount of irrelevant 'background' activities. Recent works suggested that sequential DL models (such as LSTMs) have great potential for modeling non-periodic behaviours, and in this paper we studied some LSTM training strategies for SAR. Specifically, we proposed two simple yet effective LSTM variants, i.e., a delay model and an inverse model. The delay model can effectively exploit pre-defined delay intervals as additional contextual information, while the inverse model can learn patterns from the time-series in an inverse manner, which can be complementary to the forward model (i.e., LSTM). These two LSTM variants are very practical, and they can be deemed as training strategies without alteration of the LSTM fundamentals. We also studied some additional LSTM training strategies, which can further improve the accuracy. We evaluated our models on two public datasets, and the promising results suggested the effectiveness of our approaches in recognizing sporadic activities.