EUVIMEDEUVIMED entdecken
Crossref

Wi-DMAR: Cross-Domain Human Activity Recognition via an Enhanced Conditional Diffusion Model

Caibin Tang, Pingping Tang, Hui Zhang, Jiong Jin, Shiwen Mao

IEEE Internet of Things Journal · 01.06.2026 · 13 (11) · 25343-25353

DOI: 10.1109/jiot.2026.3676817 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Internet of Things Journal, erschienen am 01.06.2026. Verfasst von Caibin Tang und 4 weiteren Autor:innen. Die Arbeit nennt 33 Literaturangaben. Gefördert von: NSFC, National Key Research and Development Program of China, NSF.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Wi-Fi-based human activity recognition (HAR) has emerged as a focal point within the Internet of Things landscape, due to its nonintrusive sensing capabilities and inherent privacy-preserving advantages. In existing Wi-Fi-based HAR research, channel state information (CSI) is primarily utilized to capture activity-related features and enable recognition. However, CSI-based cross-domain HAR remains challenged by issues, such as redundant subcarriers, limited samples in the target domain, and high sensitivity of CSI to environmental variations. To address these challenges, this article proposes Wi-DMAR, a Wi-Fi-based cross-domain HAR framework that integrates three key modules. First, an adaptive subcarrier selection module computes the correlation between each subcarrier and the principal components, identifies subcarriers with high contribution, preserves essential activity-related features, reduces data dimensionality, and lowers computational overhead. Second, a conditional diffusion-based data augmentation module employs a Transformer-based feature extractor to capture domain-specific representations of the target-domain data, and optimizes the domain consistency loss and domain-guided diffusion loss to generate pseudo-samples that resemble the target-domain distribution, thereby mitigating the sample scarcity. Third, an activity recognition module based on sample similarity learning reformulates the traditional label classification problem into a sample comparison task, by quantifying similarity between samples, and it performs the activity recognition and enhances the cross-domain generalization. Experimental results demonstrate that Wi-DMAR achieves the superior recognition accuracy compared with state-of-the-art cross-domain HAR methods, such as DiffAR and MetaAct. Ablation studies further confirm that each core component contributes positively to performance improvements.

Förderung

  • NSFC
  • National Key Research and Development Program of China
  • NSF
  • Basic Research Project of Jiangsu
  • Natural Science Foundation of Anhui Province
  • Anhui Postdoctoral Scientific Research Program Foundation

Zitieren

APA

Tang, C., Tang, P., Zhang, H., Jin, J., & Mao, S. (2026). Wi-DMAR: Cross-Domain Human Activity Recognition via an Enhanced Conditional Diffusion Model. IEEE Internet of Things Journal, 13(11), 25343-25353. https://doi.org/10.1109/jiot.2026.3676817

Vancouver

Tang C, Tang P, Zhang H, Jin J, Mao S. Wi-DMAR: Cross-Domain Human Activity Recognition via an Enhanced Conditional Diffusion Model. IEEE Internet of Things Journal. 2026;13(11):25343-25353. doi:10.1109/jiot.2026.3676817

BibTeX
@article{Tang2026,
  author = {Tang, Caibin and Tang, Pingping and Zhang, Hui and Jin, Jiong and Mao, Shiwen},
  title = {Wi-DMAR: Cross-Domain Human Activity Recognition via an Enhanced Conditional Diffusion Model},
  journal = {IEEE Internet of Things Journal},
  year = {2026},
  volume = {13},
  number = {11},
  pages = {25343--25353},
  doi = {10.1109/jiot.2026.3676817}
}
Kostenlos starten →