This paper introduces a perceptive fusion electronic skin (PFES) for gesture interaction. The PFES uses a bioinspired hierarchical structure and a magnetostrictive alloy film to sense both curvature (joint shape) and magnetism (joint position). A reinforced knowledge distillation method enables autonomous selection of knowledge relevant to user hand movements, allowing for model compression and deployment on wearable devices. The PFES fuses curvature and magnetism information for gesture recognition and haptic feedback, facilitating human-machine interaction.
