Research Article

Improving the Performance of Closet-Set Classification in Human Activity Recognition by Applying a Residual Neural Network Architecture

Daniel Rodriguez Gonzalez
Colorado Technical University, USA
Yanzhen Qu
Colorado Technical University, USA
* Corresponding author
DOI icon 10.24018/ejece.2022.6.2.433

Read Counter
411

Downloads
268

Citations

Share

Article Sidebar

Submitted 2022-03-23
Published 2022-04-15

Read counter = 411 times

Table of Contents

Article Main Content

The Time Series Classification Residual Network (TSCResNet) deep learning model is introduced in this study to improve the classification performance in the human activity recognition (HAR) problem. Specifically in the context of closed-set classification where all labels or classes are present during the model training phase. This contrasts with open-set classification where new, unseen activities are introduced to the HAR system after the training phase. The proposed TSCResNet model is evaluated with the benchmark PAMAP2 Physical Activity Monitoring Dataset. By using the same quantitative methods and data preprocessing protocols as previous research in the field of closed-set HAR, results show that the TSCResNet model architecture is able to achieve improved classification results across accuracy, the weighted F1-score, and mean F1-score.

Keywords: Closed-set classification, deep learning, human activity recognition, residual neural networks

References

  1. Buza K. Time series classification and its applications. 8th International Conference on Web Intelligence, Mining and Semantics; 2018 Jun 25-27; Novi Sad, Serbia.
     Google Scholar
  2. Lv T, Wang X, Jin, L, Xiao Y, Song, M. Margin-based deep learning networks for human activity recognition. Sensors. 2020;20(7). https://doi.org/10.3390/s20071871
     Google Scholar
  3. UCI Machine Learning Repository Irvine, CA: University of California, School of Information and Computer Science [Internet]. [Retrieved 2021 May 28]. Available from: http://archive.ics.uci.edu/ml.
     Google Scholar
  4. Reiss A, Stricker D. Creating and benchmarking a new dataset for physical activity monitoring. 5th International Conference on PErvasive Technologies Related to Assistive Environments; 2012; Crete Island, Greece.
     Google Scholar
  5. Guan Y, Pl?tz T. Ensembles of deep lstm learners for activity recognition using wearables. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 2017;1(2):1-28. https://doi.org/10.1145/3090076.
     Google Scholar
  6. Moya Rueda F, Grzeszick R, Fink G, Feldhorst S, ten Hompel M. Convolutional neural networks for human activity recognition using body-worn sensors. Informatics. 2018; 5(2). https://doi.org/10.3390/informatics5020026.
     Google Scholar
  7. Saez Y, Baldominos A, Isasi P. A comparison study of classifier algorithms for cross-person physical qctivity recognition. Sensors. 2016;17(1). https://doi.org/10.3390/s17010066.
     Google Scholar
  8. Ordonez FJ, Roggen D. Deep convolutional and lstm recurrent neural networks for multimodal wearable activity recognition. Sensors. 2016;16(1). https://doi.org/10.3390/s16010115.
     Google Scholar
  9. Li F, Shirahama K, Nisar MA, Koping L, Grzegorzek M. Comparison of feature learning methods for human activity recognition using wearable sensors. Sensors. 2018;18(2). https://doi.org/10.3390/s18020679.
     Google Scholar
  10. Panwar M, Dyuthi SR, Prakash CK, Biswas D, Acharyya A, Maharatna K, et al. CNN based approach for activity recognition using a wrist-worn accelerometer. Conf Proc IEEE Eng Med Biol Soc; 2017, 2438-2441. https://doi.org/10.1109/EMBC.2017.8037349.
     Google Scholar
  11. Gu W, Zhou Z, Zhou Y, He M, Zou H, Zhang L. Predicting blood glucose dynamics with multi-time-series deep learning. 15th ACM Conference on Embedded Network Sensor Systems. 2017;Delft, The Netherlands. https://doi.org/10.1145/3131672.3136965.
     Google Scholar
  12. Han L, Yu C, Xiao K, Zhao X. A new method of mixed gas identification based on a convolutional neural network for time series classification. Sensors. 2019;19(9). https://doi.org/10.3390/s19091960
     Google Scholar
  13. Uddin MZ, Hassan MM. Activity recognition for cognitive assistance using body sensors data and deep convolutional neural network. IEEE Sensors Journal. 2019; 19(19): 8413-8419. https://doi.org/10.1109/jsen.2018.2871203.
     Google Scholar
  14. Yang JB, Nguyen MN, San PP, Li XL, Krishnaswamy S. Deep convolutional neural networks on multichannel time series for human activity recognition. 24th International Joint Conference on Artificial Intelligence. 2015.
     Google Scholar
  15. He K, Zhang X, Ren S, Sun J. Deep residual learning for image recognition. IEEE conference on computer vision and pattern recognition. 2016: 770-778.
     Google Scholar
  16. Lines J, Bagnall A. Time series classification with ensembles of elastic distance measures. Data Mining and Knowledge Discovery. 2014; 29(3):565-592. https://doi.org/10.1007/s10618-014-0361-2.
     Google Scholar