ClapMetrics: Decoding Users’ Gender and Age Through Smartwatch Gesture Dynamics

In this presentation, we introduce ClapMetrics – an innovative and seamless method for deducing the gender and age of smartwatch users. This system ingeniously harnesses the capabilities of the smartwatch’s integrated 3D sensors — specifically, the accelerometer and gyroscope—to capture the wearer’s arm movements during a clapping action. ClapMetrics operates on the principle that each clap is a complex interplay of motion, unique to the individual, much like a fingerprint. By analyzing these movements, ClapMetrics can gather rich biometric data without any active input or effort from the user, making it a remarkably non-intrusive technology. We will delve into the specifics of how ClapMetrics processes sensor data to make its estimations. We will explore the Neural Network’s architecture, the training process, and how the model interprets the sensor data to make reliable predictions about the wearer’s gender and age.