可穿戴计算机
计算机科学
拇指
虚拟现实
计算机视觉
人工智能
机制(生物学)
航空航天
可穿戴技术
成交(房地产)
匹配移动
有限元法
运动捕捉
Boosting(机器学习)
模拟
运动控制
虚拟机
材料科学
工作区
跟踪(教育)
虚拟演员
运动(物理)
刚度(电磁)
方向(向量空间)
作者
Lu Liu,Zengbai Ouyang,Mengze Zhang,Jiantao Yao,Xiaoli Jia,Peng Zhang,Bo Han,Yu Gu,Yundou Xu
摘要
ABSTRACT Wearable human–machine interfaces (HMIs) require sensing systems that combine high accuracy, mechanical compliance, and low system complexity. Here, we present a starfish‐inspired, hierarchically decoupled multimodal flexible sensor for high‐fidelity thumb motion perception and virtual human–machine interaction. Inspired by the distributed sensing mechanism of starfish tube feet, the architecture integrates radially distributed strain channels with spatially isolated pressure nodes. This design enables simultaneous sensing of global deformation and localized tactile stimuli while minimizing mechanical coupling. Finite element simulations and experimental characterization demonstrate effective strain amplification, low interchannel coupling, stable electromechanical performance, and long‐term durability. Combined with a Light Gradient Boosting Machine (LightGBM) classifier, the system achieves validation and test accuracies of 99.82% and 99.27%, respectively, for six thumb motions, with strong cross‐subject generalization. The decoded motions enable intuitive, real‐time control in a virtual environment without bulky sensor gloves or external tracking systems. This work establishes a bioinspired strategy for structurally decoupled multimodal flexible sensing and provides a promising platform for wearable HMIs and intelligent robotics.
科研通智能强力驱动
Strongly Powered by AbleSci AI