Large-Scale Nanofiber Manufacturing| Machine-Learning Enabled Biocompatible Capacitive-Electromyographic Bimodal Flexible Sensor for FacialExpression Recognition

Prof. Yang Li (Shandong University) & Assoc. Prof. Hongsen Niu (Jinan University): Dual-mode Sensor for Facial Expression Recognition

Challenge: Single-mode sensors suffer from poor robustness and insufficient data features in facial expression recognition. Fusing multi-sensor signals is therefore crucial to enhance the accuracy of facial expression recognition systems.

Approach: A research team led by Assoc. Prof. Hongsen Niu (Jinan University), Assoc. Prof. Yuanyue Li (Qingdao University), and Prof. Yang Li (Shandong University) innovatively developed a multi-sensor signal fusion method to capture richer facial expression features, significantly improving recognition accuracy.

Innovation 1: The team engineered a dual-coupled microstructure between electrode and dielectric layers by introducing a porous PVDF membrane and an SK-structured PVA/BMMICl membrane. BMMICl doping further induced an EDL effect, substantially enhancing the capacitive sensor's responsiveness.

Innovation 2: The dry-electrode sensor for electrophysiological monitoring eliminates skin irritation and dehydration issues associated with hydrogel electrodes, demonstrating superior long-term usability. Cytotoxicity and antibacterial tests confirmed the CEDS' excellent biocompatibility and antimicrobial properties.

Innovation 3: By integrating capacitive and electrophysiological monitoring in CEDS with a 1D-CNN system, the team achieved breakthrough recognition accuracy, showcasing the immense potential of flexible sensor technology in practical facial expression monitoring applications.