摘要
由一名新闻记者兼机器人与机器学习每日新闻编辑-调查人员讨论机器学习的新发现。根据NewsRx记者从联合王国曼彻斯特发回的新闻报道，研究表明，“机器人手的自主操作可以从触觉传感中受益，因为它可以收集关于施加力和表面特性变化的信息。本文介绍了一种安装在机器人手指上的电容式触觉传感器。”本研究经费来自北京塔山科技股份有限公司，新闻编辑引用曼奇斯特大学的研究报告：“由于其独特的结构和对材料介电常数的高灵敏度，该传感系统在机器人手指接触物体和接触物体时都能获得电容数据。”通过主成分分析(PCA)、独立成分分析(ICA)和多维尺度(MDS),将一个数据集转换为二维,然后输入两种监督分类算法,即k近邻(KNNs)和支持向量机(SVMs)。与以往的研究相比,基于MDS的SVM具有较高的材料识别精度,对塑料、塑料等3种不同材料的识别率高达98%。此外，该算法对干塑料、带水滴塑料、纸张、干玻璃和带水滴玻璃五种不同材料的识别效果良好，识别准确率高达93%，将降维方法与分类算法结合起来，计算时间可缩短60%左右。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – Investigators discuss new findings in Machine Learning. According to news reporting originating from Manchester, Unite d Kingdom, by NewsRx correspondents, research stated, “Autonomous manipulation u sing robot hands can benefit from tactile sensing, as it can collect information on variations in applied force and surface properties. This article presents a capacitive tactile sensor placed on the robot’s hand fingers.” Financial support for this research came from Beijing Tashan Technology Corporat ion Ltd. Our news editors obtained a quote from the research from the University of Manch ester, “Due to its unique structure and high sensitivity to material permittivit y, this sensing system can obtain capacitive data both when a robot finger is ap proaching an object and when it has touched the object. With threedimensional r eduction methods, that is, principal component analysis (PCA), independent compo nent analysis (ICA), and multidimensional scaling (MDS), a dataset is transforme d to be 2-D and then fed into two supervised classifications algorithms, that is , k-nearest neighbors (KNNs) and support vector machines (SVMs). In comparison t o previous studies, the MDS-based SVM achieves high material recognition accurac y, up to 98% for recognition of three different material classes, that is, plastic, paper, and glass using capacitance data only. Furthermore, it performs well in recognition of five different materials, that is, dry plastic, plastic with water drops, paper, dry glass, and glass with water drops. The reco gnition accuracy is as high as 93%. Computational time can be reduc ed by about 60% by combining the dimension reduction methods with classification algorithms.”
NETL
NSTL