ᯓ➤ ResponsiveSNNs-FallDetection

Improving Responsiveness of Fall Detection using Spiking Neural Networks
Published in the 2025 IEEE International Conference on Pervasive Computing and Communications Workshops and other Affiliated Events (PerCom Workshops)

📄 Paper Link

---

🔍 Abstract

Fall detection systems, essential for the safety of elderly individuals, have increasingly incorporated Deep Neural Networks (DNNs) for improved accuracy. However, real-time processing—especially on resource-constrained wearable devices—remains a challenge due to the computational demands of DNNs.

Recently, Spiking Neural Networks (SNNs) have shown promise for improving energy efficiency. Yet, their potential to leverage temporal dynamics for faster responsiveness has not been fully explored.

This work proposes:

• Quick Spike Encoding (QSE): An encoding method that prioritizes critical inputs based on amplitude over time.
• Linear Weighted Mean Squared Error Count (LW-MSEC) Loss Function: A loss function that emphasizes early detection by penalizing early-stage errors more heavily.

Evaluations on two fall detection datasets show significant improvements:

• 91% accuracy in just 10 time steps on SisFall (vs 25 steps in standard baselines)
• Over 60% improvement in responsiveness

---

📄 Citation

If you use this code in your research, please cite:

@INPROCEEDINGS{11038662,
  author={Sabbella, Hemanth and Mukherjee, Archit and Chuang, Tan Jeck and Yee Low, Hong and Ma, Dong and Misra, Archan},
  booktitle={2025 IEEE International Conference on Pervasive Computing and Communications Workshops and other Affiliated Events (PerCom Workshops)}, 
  title={Improving Responsiveness of Fall Detection using Spiking Neural Networks}, 
  year={2025},
  pages={98--103},
  doi={10.1109/PerComWorkshops65533.2025.00048},
  keywords={Accuracy;Conferences;Spiking neural networks;Encoding;Real-time systems;Timing;Safety;Fall detection;Wearable devices;Older adults;Spiking Neural Networks;Responsiveness;Spike Encoding;Weighted Loss Function}
}