Ultrathin asymmetric ZnFe-LDH/PDMS flexible sensor for accurate gesture recognition and high-stability human motion monitoring via geometry-embedded algorithm

The urgent need for advanced human-motion sensing in sports and healthcare drives the development of flexible piezoelectric sensors. This study introduces a high-performance, potentially eco-friendly sensor using ultrathin asymmetric ZnFe-LDH nanosheets and PDMS, exhibiting exceptional sensitivity with a pressure sensitivity of 2.316 mV/kPa and an ultra-low detection limit of 0.001 N, along with robust cycling stability (>10,000 cycles). It enables real-time monitoring of subtle motions like nodding and joint bending, and facilitates distinguishable detection of gait and hand gestures via insole and gesture recognition systems. A geometry-driven approach with hyperspherical embedding algorithm is employed for data processing, where multi-granularity temporal convolutions extract features mapped onto a hypersphere for angular similarity-based classification. This achieves 100% gesture and gait recognition accuracy via Geometry-Embedded algorithm. This work provides a low-cost, sustainable material platform and an algorithmic solution for flexible piezoelectric sensing in health monitoring and human-computer interaction, highlighting the transformative potential of ZnFe-LDH/PDMS composites for next-generation flexible electronics.