Humanoids learn touch modalities identification via multi-modal robotic skin and robust tactile descriptors
Humanoids learn touch modalities identification via multi-modal robotic skin and robust tactile descriptors
复制标题
DOI：
10.1080/01691864.2015.1095652
复制
发表时间：
2015-11-02
影响因子：
2
通讯作者：
Cheng, Gordon
中科院分区：
文献类型：
--
作者：
Kaboli, Mohsen;Long, Alex;Cheng, Gordon
在本文中，我们提出了一种通过人形机器人的触觉感知来识别触摸形态的新方法。在这方面，我们为 NAO 人形机器人配备了整个上半身覆盖的多模态人造皮肤。我们提出了一组受生物学启发的特征描述符，以提供用于触摸分类的稳健且抽象的触觉信息。这些功能被证明对于人形机器人的接触位置和移动是不变的，并且能够处理单点和多点触摸动作。为了对我们的方法进行比较，重新实现和评估了现有的方法。实验结果表明，使用所提出的特征描述符和 SVM 分类器，类人机器人可以区分不同的单点触摸模式，识别率为 96.79%。此外，它还可以识别多个触摸动作，识别率高达93.03%。
In this paper, we present a novel approach for touch modality identification via tactile sensing on a humanoid. In this respect, we equipped a NAO humanoid with whole upper body coverage of multi-modal artificial skin. We propose a set of biologically inspired feature descriptors to provide robust and abstract tactile information for use in touch classification. These features are demonstrated to be invariant to location of contact and movement of the humanoid, as well as capable of processing single and multi-touch actions. To provide a comparison of our method, existing approaches were reimplemented and evaluated. The experimental results show that the humanoid can distinguish different single touch modalities with a recognition rate of 96.79% while using the proposed feature descriptors and SVM classifier. Furthermore, it can recognize multiple touch actions with 93.03% recognition rate.