Real-Time Assault Detection Using Wearable Sensors

What if a wearable device could recognize danger as it unfolds?

In this project, we explore how multimodal signals can be leveraged to identify patterns associated with stress, fear, and potential assault scenarios. By combining sensor fusion with advanced machine learning, we aim to move beyond activity tracking toward proactive personal safety systems. 

Our goal is to design a real-time wrist-worn assault detection system that can automatically initiate a response even when individuals are unable to take action due to physiological freeze or physical restraint.