A high-performance flexible coaxial BTO/PAN piezoelectric nanofibers for facial expression analysis

Piezoelectric nanofibers that combine the flexibility of organic polymers with the high piezoelectricity of inorganic ceramics have garnered significant attention in the field of wearable devices. However, a significant obstacle is the loss of sensing ability, due to the weak interfacial bonding between inorganic fillers and organic matrices. Here, a novel composite nanofiber is proposed via coaxial electrospinning strategy, which couples two functional units of Barium Titanate (BTO)/Polyacrylonitrile (PAN) into the entirety. In addition, the dual-solvent synergistic strategy combined with surface modification enhances interfacial bonding, fully leveraging the synergistically enhanced piezoelectricity to boost performance. The developed piezoelectric sensor, possessing high sensitivity and durability, can accurately distinguish various subtle facial muscle movements. More importantly, with the aid of the optimized deep learning model based on the electrical outputs, the different emotional states, including surprise, disgust, happiness, anger, and sadness, can be classified with an accuracy of 98.70 %. This capability of the sensor and related algorithm model can underscore the practical utility of human physiological monitoring, which provides an opens new avenues for the development of human–machine interaction technologies.