Liquid Metal‐Based Eutectohydrogel Strain Sensor for Ultra‐Low Hysteresis Human–Machine Interaction and High‐Precision Sign Language Recognition

ABSTRACT Flexible strain sensors for human‐computer interaction and embodied intelligence require stable electrical output under continuous deformation to enable real‐time gesture control for robots and virtual scenes. However, hydrogel‐based sensors are often subject to viscoelastic dissipation, environmental fluctuations, and structural instability, which can lead to signal drift. To address this challenge, this study ingeniously combined tannic acid (TA)‐mediated liquid metal (LM) nanocomposites combined with polymerizable deep eutectic solvents (PDES) to successfully fabricate a conductive eutectohydrogel through the synergistic effect of stable conductive phases and controllable microenvironments. The resulting conductive eutectohydrogels exhibit high stretchability (∼430%), good electrical conductivity (1.4 S m −1 ) and self‐healing properties. Notably, the corresponding strain sensor provides ultra‐low hysteresis (1.7%) and long‐term stability, allowing high fidelity and reproducible detection of fine deformations. To demonstrate the sensor's practicality, it was integrated into a flexible glove for human‐computer interaction and sign language recognition, as well as for machine learning‐assisted signal analysis. The system achieved a recognition accuracy 97.4% for gesture classification and 99.7% for virtual character control. These conductive eutectohydrogels for high‐precision sensing show promise for wearable smart devices and offer new avenues for immersive human–machine interfaces and soft robotic systems.