35
Health & Science3h 12m ago

A unified framework combining transformer-based temporal modeling, lightweight architectures, and federated learning for privacy-preserving gait classification and anomaly detection in healthcare surveillance was proposed and evaluated.

Archive Window: 7 Days Left

Controlled federated learning environment

Who
Researchers
What
A unified framework combining transformer-based temporal modeling, lightweight architectures, and federated learning for privacy-preserving gait classification and anomaly detection in healthcare surveillance was proposed and evaluated.
When
Fri, 17 Jul 2026 21:37:16 GMT · 3h 12m ago
Where
Controlled federated learning environment ·
Why
To address limitations of existing methods in model complexity, computational requirements, and privacy concerns for healthcare surveillance and clinical gait analysis.
The Frontline Impact

How this affects you

The proposed framework offers a privacy-preserving approach for healthcare monitoring by enabling efficient gait analysis on edge devices. MobileViT-Small demonstrated suitability for edge deployment with 94.0% accuracy.

Story chain

1 event in this thread
No related history yet - this is the origin event.

Verified Sources & Citations