Domain adaptation methods for lab-to-field human context recognition

dc.contributor.authorAlajaji, Abdulaziz
dc.contributor.authorGerych, Walter
dc.contributor.authorBuquicchio, Luke
dc.contributor.authorChandrasekaran, Kavin
dc.contributor.authorMansoor, Hamid
dc.contributor.authorAgu, Emmanuel
dc.contributor.authorRundensteiner, Elke
dc.date.accessioned2024-02-02T19:35:05Z
dc.date.available2024-02-02T19:35:05Z
dc.date.copyright2023en_US
dc.date.issued2023
dc.description.abstractHuman context recognition (HCR) using sensor data is a crucial task in Context-Aware (CA) applications in domains such as healthcare and security. Supervised machine learning HCR models are trained using smartphone HCR datasets that are scripted or gathered in-the-wild. Scripted datasets are most accurate because of their consistent visit patterns. Supervised machine learning HCR models perform well on scripted datasets but poorly on realistic data. In-the-wild datasets are more realistic, but cause HCR models to perform worse due to data imbalance, missing or incorrect labels, and a wide variety of phone placements and device types. Lab-to-field approaches learn a robust data representation from a scripted, high-fidelity dataset, which is then used for enhancing performance on a noisy, in-the-wild dataset with similar labels. This research introduces Triplet-based Domain Adaptation for Context REcognition (Triple-DARE), a lab-to-field neural network method that combines three unique loss functions to enhance intra-class compactness and inter-class separation within the embedding space of multi-labeled datasets: (1) domain alignment loss in order to learn domain-invariant embeddings; (2) classification loss to preserve task-discriminative features; and (3) joint fusion triplet loss. Rigorous evaluations showed that Triple-DARE achieved 6.3% and 4.5% higher F1-score and classification, respectively, than state-of-the-art HCR baselines and outperformed non-adaptive HCR models by 44.6% and 10.7%, respectively.en_US
dc.description.reviewstatusRevieweden_US
dc.description.scholarlevelFacultyen_US
dc.description.sponsorshipThis work was supported in part by the Computer Science Department at Worcester Polytechnic Institute and the DARPA WASH under Grant HR00111780032-WASH-FP-031 and in part by DARPA under Agreement FA8750-18-2-0077. The U.S. Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright notation thereon. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of DARPA or the U.S. Government.en_US
dc.identifier.citationAlajaji, A., Gerych, W., Buquicchio, L., Chandrasekaran, K., Mansoor, H., Agu, E., & Rundensteiner, E. (2023). Domain adaptation methods for lab-to-field human context recognition. Sensors, 23(6), 3081. https://doi.org/10.3390/s23063081en_US
dc.identifier.urihttps://doi.org/10.3390/s23063081
dc.identifier.urihttp://hdl.handle.net/1828/15918
dc.language.isoenen_US
dc.publisherSensorsen_US
dc.subjectubiquitous computing
dc.subjectdomain adaptation
dc.subjectcontext aware systems
dc.subjectmachine learning
dc.subject.departmentDepartment of Computer Science
dc.titleDomain adaptation methods for lab-to-field human context recognitionen_US
dc.typeArticleen_US

Files

Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
alajaji_abdulaziz_sensors_2023.pdf
Size:
1.66 MB
Format:
Adobe Portable Document Format
Description:
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
2 KB
Format:
Item-specific license agreed upon to submission
Description: