A Rapid Adaptation Approach for Dynamic Air-Writing Recognition Using Wearable Wristbands with Self-Supervised Contrastive Learning

• Utilizes self-supervised contrastive learning to reduce dependency on labeled data, enabling rapid adaptation with few-shot fine-tuning. • Wearable wristband with four-channel ionic hydrogel sensor array achieves high-sensitivity capacitance output for dynamic gesture tracking. • Achieves 94.9% accuracy in diverse scenarios including eight-direction commands and air-writing of numbers and letters. • Demonstrates practical utility in human–machine interaction applications such as game control, calculators, and multilingual login systems.