EUVIMEDOpdag EUVIMED
Open access · CC BY 4.0

Generalizable Fall Detection via Fall-Specific Adaptation of a Skeleton Action Foundation Model

Sehyeon Jang, Minyoung Kim, Chanyoung Lee, Chaerin Lee, Soohwan Song

Electronics · 2026-11-18 · 15 (22) · 5376

DOI: 10.3390/electronics15225376 ↗

At a glance

Journal article in Electronics, published online on 2026-11-18. Written by Sehyeon Jang and 4 co-authors. 29 references. The full text is freely available (open access, CC BY 4.0). Funded by: Ministry of Science and ICT, Korea Institute of Energy Technology Evaluation and Planning.

AbstractBibliografiske oplysningerCiteRelated studies

Abstract (original language)

Vision-based fall detection remains challenging when falls resemble routine daily activities and test environments differ from training conditions. We propose a fall-specific adaptation framework that extends a pretrained skeleton-based action recognizer while preserving its original pathway for classifying activities of daily living (ADL). RGB videos are converted into 3D skeleton sequences and encoded by a frozen Dense Spatio-Temporal Encoder (DSTE). A lightweight residual adapter refines the extracted features to distinguish among falls, transitions into a lying posture, lying postures, and other activities. To combine complementary supervision, we jointly train the adapter on SAFER-Activities and FU-Kinect-Fall using separate classification heads for four-class window-level and binary clip-level supervision, respectively. We then freeze the adapter and train a dedicated four-class classifier while leaving the original ADL pathway unchanged. Across five training seeds, the method achieves video-level F1-scores of 89.13±1.70% and 91.50±0.88% (mean ± standard deviation) on the unseen CAUCAFall and Le2i datasets, respectively. The corresponding mean improvements over a classifier trained on unadapted DSTE features are 2.98 and 0.20 percentage points. The method also achieves higher mean F1-scores than the dedicated fall detectors evaluated in this study. The framework provides a lightweight fall-specific extension to a pretrained skeleton-based ADL recognizer, producing four-class fall-related predictions alongside the unchanged original ADL classification pathway.

Funding

  • Ministry of Science and ICT
  • Korea Institute of Energy Technology Evaluation and Planning

Cite

APA

Jang, S., Kim, M., Lee, C., Lee, C., & Song, S. (2026). Generalizable Fall Detection via Fall-Specific Adaptation of a Skeleton Action Foundation Model. Electronics, 15(22), 5376. https://doi.org/10.3390/electronics15225376

Vancouver

Jang S, Kim M, Lee C, Lee C, Song S. Generalizable Fall Detection via Fall-Specific Adaptation of a Skeleton Action Foundation Model. Electronics. 2026;15(22):5376. doi:10.3390/electronics15225376

BibTeX
@article{Jang2026,
  author = {Jang, Sehyeon and Kim, Minyoung and Lee, Chanyoung and Lee, Chaerin and Song, Soohwan},
  title = {Generalizable Fall Detection via Fall-Specific Adaptation of a Skeleton Action Foundation Model},
  journal = {Electronics},
  year = {2026},
  volume = {15},
  number = {22},
  pages = {5376},
  doi = {10.3390/electronics15225376}
}
Start for free →