Comparison of Feature Learning Methods for Human Activity Recognition Using Wearable Sensors

• Published in: Sensors (Basel, Switzerland) Vol. 18; no. 2; p. 679
• Main Authors: Li, Frédéric, Shirahama, Kimiaki, Nisar, Muhammad, Köping, Lukas, Grzegorzek, Marcin
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 24.02.2018; MDPI
• Subjects: Classification; Datasets; deep neural networks; evaluation framework; feature learning; human activity recognition; Image retrieval; multimodal time series processing; Neural networks; Neurons; Sensors; Support vector machines; multimodal time series processing; feature learning; human activity recognition; deep neural networks; evaluation framework
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s18020679

Getting a good feature representation of data is paramount for Human Activity Recognition (HAR) using wearable sensors. An increasing number of feature learning approaches—in particular deep-learning based—have been proposed to extract an effective feature representation by analyzing large amounts of data. However, getting an objective interpretation of their performances faces two problems: the lack of a baseline evaluation setup, which makes a strict comparison between them impossible, and the insufficiency of implementation details, which can hinder their use. In this paper, we attempt to address both issues: we firstly propose an evaluation framework allowing a rigorous comparison of features extracted by different methods, and use it to carry out extensive experiments with state-of-the-art feature learning approaches. We then provide all the codes and implementation details to make both the reproduction of the results reported in this paper and the re-use of our framework easier for other researchers. Our studies carried out on the OPPORTUNITY and UniMiB-SHAR datasets highlight the effectiveness of hybrid deep-learning architectures involving convolutional and Long-Short-Term-Memory (LSTM) to obtain features characterising both short- and long-term time dependencies in the data.