Sulfonated lignin-doped dynamic coordination hydrogel as flexible sensor for intelligent gesture recognition via machine learning.

The translation of subtle human motions into precise digital commands via wearable electronics is often hindered by material fatigue and signal noise, necessitating advanced hydrogel engineering and intelligent data interpretation. In this study, a multifunctional conductive hydrogel was successfully prepared through Fe3+ coordination crosslinking and sulfonated lignin doping. The as-optimized hydrogel features a highly interconnected porous structure and dynamic crosslinking network, demonstrating outstanding mechanical properties and stable sensing performance. The flexible sensor can be attached to the human body to realize real-time monitoring of joint movements, swallowing, blinking, and frowning, with excellent skin conformability and signal stability. Furthermore, digit gestures (0-9) were designed based on standard Morse code, and a two-layer long short-term memory (LSTM) network, which is a kind of machine learning model, was constructed for gesture recognition. After three-fold data augmentation, the model achieved 98.89% recognition accuracy on the test set. This integrated strategy provides a new technical solution for flexible wearable sensors in health monitoring, human-machine interaction, and silent communication.