Deep, Convolutional, and Recurrent Models for Human Activity Recognition using Wearables

TLDR

Human activity recognition is increasingly adopting deep learning, yet a comprehensive comparison of deep, convolutional, and recurrent models across diverse tasks remains lacking. This study rigorously evaluates deep, convolutional, and recurrent approaches on three wearable‑sensor datasets to determine their suitability for various HAR tasks. We train recurrent models with a novel regularization scheme, conduct thousands of experiments with random configurations, analyze hyperparameter effects via fANOVA, and provide practical guidelines for applying deep learning in HAR. The recurrent models with the new regularization outperform the state‑of‑the‑art on a large benchmark dataset.

Abstract

Human activity recognition (HAR) in ubiquitous computing is beginning to adopt deep learning to substitute for well-established analysis techniques that rely on hand-crafted feature extraction and classification techniques. From these isolated applications of custom deep architectures it is, however, difficult to gain an overview of their suitability for problems ranging from the recognition of manipulative gestures to the segmentation and identification of physical activities like running or ascending stairs. In this paper we rigorously explore deep, convolutional, and recurrent approaches across three representative datasets that contain movement data captured with wearable sensors. We describe how to train recurrent approaches in this setting, introduce a novel regularisation approach, and illustrate how they outperform the state-of-the-art on a large benchmark dataset. Across thousands of recognition experiments with randomly sampled model configurations we investigate the suitability of each model for different tasks in HAR, explore the impact of hyperparameters using the fANOVA framework, and provide guidelines for the practitioner who wants to apply deep learning in their problem setting.