Scaling Human Activity Recognition via Deep Learning-based Domain Adaptation
| dc.contributor.author | Khan, Md Abdullah Al Hafiz | |
| dc.contributor.author | Roy, Nirmalya | |
| dc.contributor.author | Misra, Archan | |
| dc.date.accessioned | 2018-09-04T19:25:06Z | |
| dc.date.available | 2018-09-04T19:25:06Z | |
| dc.date.issued | 2018-08-23 | |
| dc.description | © 2018 IEEE; 2018 IEEE International Conference on Pervasive Computing and Communications (PerCom) | en_US |
| dc.description.abstract | We investigate the problem of making human activity recognition (AR) scalable-i.e., allowing AR classifiers trained in one context to be readily adapted to a different contextual domain. This is important because AR technologies can achieve high accuracy if the classifiers are trained for a specific individual or device, but show significant degradation when the same classifier is applied context-e.g., to a different device located at a different on-body position. To allow such adaptation without requiring the onerous step of collecting large volumes of labeled training data in the target domain, we proposed a transductive transfer learning model that is specifically tuned to the properties of convolutional neural networks (CNNs). Our model, called HDCNN, assumes that the relative distribution of weights in the different CNN layers will remain invariant, as long as the set of activities being monitored does not change. Evaluation on real-world data shows that HDCNN is able to achieve high accuracy even without any labeled training data in the target domain, and offers even higher accuracy (significantly outperforming competitive shallow and deep classifiers) when even a modest amount of labeled training data is available. | en_US |
| dc.description.sponsorship | The authors thank the shepherd Stephan Sigg and anonymous reviewers for their constructive feedback and comments. This research is partially supported by the ONR under grant N00014-15-1-2229, and partially by the Singapore Ministry of Education Academic Research Fund Tier2 under research grant MOE2014-T2-1063. | en_US |
| dc.description.uri | http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8444585&isnumber=8444570 | en_US |
| dc.format.extent | 9 PAGES | en_US |
| dc.genre | conference papers and proceedings preprints | en_US |
| dc.identifier | doi:10.13016/M2QV3C70M | |
| dc.identifier.citation | M. A. A. H. Khan, N. Roy and A. Misra, "Scaling Human Activity Recognition via Deep Learning-based Domain Adaptation," 2018 IEEE International Conference on Pervasive Computing and Communications (PerCom), Athens, Greece, 2018, pp. 1-9. | en_US |
| dc.identifier.uri | 10.1109/PERCOM.2018.8444585 | |
| dc.identifier.uri | http://hdl.handle.net/11603/11220 | |
| dc.language.iso | en_US | en_US |
| dc.publisher | IEEE | en_US |
| dc.relation.isAvailableAt | The University of Maryland, Baltimore County (UMBC) | |
| dc.relation.ispartof | UMBC Information Systems Department Collection | |
| dc.relation.ispartof | UMBC Faculty Collection | |
| dc.relation.ispartof | UMBC Student Collection | |
| dc.relation.ispartof | UMBC Faculty Collection | |
| dc.relation.ispartof | UMBC Student Collection | |
| dc.rights | This item may be protected under Title 17 of the U.S. Copyright Law. It is made available by UMBC for non-commercial research and education. For permission to publish or reproduce, please contact the author. | |
| dc.subject | Adaptation models | en_US |
| dc.subject | Activity recognition | en_US |
| dc.subject | Data models | en_US |
| dc.subject | Training data | en_US |
| dc.subject | Feature extraction | en_US |
| dc.subject | Accelerometers | en_US |
| dc.subject | Mobile Pervasive & Sensor Computing Lab | en_US |
| dc.title | Scaling Human Activity Recognition via Deep Learning-based Domain Adaptation | en_US |
| dc.type | Text | en_US |