首页> 外国专利> Method and device for authenticating signatures and e-learning methods and comparing them with previously stored representations

Method and device for authenticating signatures and e-learning methods and comparing them with previously stored representations

机译:用于验证签名和电子学习方法并将它们与先前存储的表示进行比较的方法和设备

摘要

"METHOD AND APPARATUS FOR AUTHENTICATING SIGNATURES AND METHODS OF ELECTRONIC LEARNING AND COMPARISON OF THE SAME WITH PREVIOUSLY STORED REPRESENTATIONS". Signature authentication method including the steps of supplying the signature sample and storing the representative signature data, converting the data into high-dimensional vectors, feeding the high-dimensional vectors to an unsupervised neural network, carrying out an extraction process of main components of high order on the vectors of high dimensions, in order to, thus, identify the groupings of high dimension points, and analysis of the groupings of high dimension points, to determine, based on previously stored information, the authenticity of the signature . Also a device for this authentication, including a signature sample collection device and storage of the representative signature data, a conversion device connected downstream to the sample collection device for converting the data into large dimension vectors, a neural network unsupervised to receive the high dimensions and carry out a process of extraction of main components of high order on the vectors of high dimensions, in order to, thus, identify the clusters of high dimension points, and an analysis device connected to the unsupervised neural network, to analyze the groupings of large points in order to determine the authenticity of the signature.
机译:“用于认证签名的方法和装置以及电子学习的方法和具有相同存储代表的方法的比较”。签名认证方法包括以下步骤:提供签名样本并存储代表性签名数据,将数据转换为高维向量,将高维向量馈送到无监督神经网络,执行高阶主要成分的提取过程为了识别高维点的分组,并分析高维点的分组,以便基于先前存储的信息,确定签名的真实性,从而确定高维点的分组。也是用于该认证的设备,包括签名样本收集设备和代表性签名数据的存储,在下游连接到样本收集设备以将数据转换为大维向量的转换设备,不受监督以接收高维的神经网络,以及在高维向量上进行高阶主要成分的提取过程,从而识别高维点的簇,并连接到无监督神经网络的分析设备来分析大维的分组点,以确定签名的真实性。

著录项

  • 公开/公告号BR0007250A

    专利类型

  • 公开/公告日2002-10-15

    原文格式PDF

  • 申请/专利权人 COMPUTER ASSOCIATES THINK INC;

    申请/专利号BR20000007250

  • 发明设计人 EYTAN SUCHARD;YOSSI ANVI;

    申请日2000-01-13

  • 分类号G06K9/00;

  • 国家 BR

  • 入库时间 2022-08-22 00:46:16

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号