CSI Human Activity Dataset
Source: Research Poster

An Ensemble Framework for Fall Detection

Using Multivariate Wi-Fi Channel State Information (CSI)

Read the paper Explore the code
Background Image
Source: Secure Safety Solutions

Background

  • Falls are a significant global health concern, especially for the elderly population.
  • Traditional fall detection systems can be costly, intrusive, and inaccurate.
  • Wi-Fi sensing offers a non-invasive, cost-effective alternative using existing infrastructure.

Rationale

  • The goal was to improve the effectiveness of fall detection systems using Wi-Fi sensing.
  • This project focused on developing a signal processing framework tailored for this purpose.
  • Key techniques include filtering, feature extraction, and predictive algorithms to optimize performance.
Rationale Image
Source: Wireless Sensing and Communication Lab
Conclusions Image
Source: Research Poster

Conclusions

  • The developed framework outperforms previous Wi-Fi-based fall detection systems.
  • MANOVA results showed a significant improvement with a p-value of 0.022.
  • Wi-Fi sensing is positioned as a promising alternative for future fall detection advancements.