Enhancing Wearable Fall Detection System via Synthetic Data

This study demonstrates how synthetic data generation using diffusion models can mitigate data imbalance in fall detection systems. It highlights improved accuracy and robustness across multiple activity classes, enhancing the reliability of wearable-based fall detection.

Recommended citation: Debnath, Minakshi, Sana Alamgeer, Md Shahriar Kabir, and Anne H. H. Ngu. (2025). "Enhancing Wearable Fall Detection System via Synthetic Data." Sensors, 25(15):4639.
Paper Link