Towards a Robust WiFi-based Fall Detection with Adversarial Data Augmentation

TL;DR

This paper explores adversarial data augmentation to improve the robustness of WiFi-based fall detection systems, highlighting modest gains in generalization across unseen environments.

Contribution

It introduces a method of adversarial data augmentation aimed at enhancing the robustness of WiFi-based fall detection models in unseen domains.

Findings

Slight improvement in unseen domain performance

Adversarial augmentation provides some generalization benefits

Performance gains are not statistically significant

Abstract

Recent WiFi-based fall detection systems have drawn much attention due to their advantages over other sensory systems. Various implementations have achieved impressive progress in performance, thanks to machine learning and deep learning techniques. However, many of such high accuracy systems have low reliability as they fail to achieve robustness in unseen environments. To address that, this paper investigates a method of generalization through adversarial data augmentation. Our results show a slight improvement in deep learning-systems in unseen domains, though the performance is not significant.