For a better experience, click the Compatibility Mode icon above to turn off Compatibility Mode, which is only for viewing older websites.

Gesture Control Cuts Parameters for Wearable Tech

wearable gesture recognition

Quantum Zeitgeist, February 17, 2026

Researchers are tackling the significant challenge of enabling gesture recognition in wearable e-textiles, devices severely limited by power, computational resources, and size. Daniel Schwartz, Dario Salvucci, and Yusuf Osmanlioglu, all from Drexel University, alongside Richard Vallett, Genevieve Dion, and Ali Shokoufandeh, also from Drexel University, present a novel approach utilising a convexified attention mechanism to dramatically reduce computational demands. This work is particularly significant as it achieves 100% accuracy in recognising both tap and swipe gestures on a textile-based sensor, while requiring only 120-360 parameters, a 97% reduction compared to existing methods, and boasting sub-millisecond inference times.

View Story