Skip to main content

Research Repository

Advanced Search

Personalised human activity recognition using matching networks.

Sani, Sadiq; Wiratunga, Nirmalie; Massie, Stewart; Cooper, Kay

Authors

Sadiq Sani



Contributors

Michael T. Cox
Editor

Peter Funk
Editor

Shahina Begum
Editor

Abstract

Human Activity Recognition (HAR) is typically modelled as a classification task where sensor data associated with activity labels are used to train a classifier to recognise future occurrences of these activities. An important consideration when training HAR models is whether to use training data from a general population (subject-independent), or personalised training data from the target user (subject-dependent). Previous evaluations have shown personalised training to be more accurate because of the ability of resulting models to better capture individual users' activity patterns. From a practical perspective however, collecting sufficient training data from end users may not be feasible. This has made using subject-independent training far more common in real-world HAR systems. In this paper, we introduce a novel approach to personalised HAR using a neural network architecture called a matching network. Matching networks perform nearest-neighbour classification by reusing the class label of the most similar instances in a provided support set, which makes them very relevant to case-based reasoning. A key advantage of matching networks is that they use metric learning to produce feature embeddings or representations that maximise classification accuracy, given a chosen similarity metric. Evaluations show our approach to substantially out perform general subject-independent models by at least 6% macro-averaged F1 score.

Citation

SANI, S., WIRATUNGA, N., MASSIE, S. and COOPER, K. 2018. Personalised human activity recognition using matching networks. In Cox, M.T., Funk, P. and Begum, S. (eds.) Case-based reasoning research and development: proceedings of the 26th International conference on case-based reasoning (ICCBR 2018), 9-12 July 2018, Stockholm, Sweden. Lecture notes in computer science, 11156. Cham: Springer [online], pages 339-353. Available from: https://doi.org/10.1007/978-3-030-01081-2_23

Conference Name 26th International conference on case-based reasoning (ICCBR 2018)
Conference Location Stockholm, Sweden
Start Date Jul 9, 2018
End Date Jul 12, 2018
Acceptance Date May 21, 2018
Online Publication Date Oct 9, 2018
Publication Date Nov 8, 2018
Deposit Date Jun 5, 2018
Publicly Available Date Oct 9, 2018
Print ISSN 0302-9743
Electronic ISSN 1611-3349
Publisher Springer
Pages 339-353
Series Title Lecture notes in computer science
Series Number 11156
Series ISSN 1611-3349
ISBN 9783030010805
DOI https://doi.org/10.1007/978-3-030-01081-2_23
Keywords Human activity recognition; Matching network; Casebased reasoning; Matching networks
Public URL http://hdl.handle.net/10059/2944