• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International conference on artificial intelligence >Determining Signal Source Integrity Using a Semi-supervised Pattern Classification System
【24h】

Determining Signal Source Integrity Using a Semi-supervised Pattern Classification System

机译：使用半监督模式分类系统确定信号源完整性

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

This paper focuses on a unique signal classification problem that employs digital signal processing techniques to first, separate 57 audio signals into two signal-type categories (A and B) and second, to further classify the integrity of the category A signal sources. Short-Time Fourier Transform and central tendency analyses are employed to distinguish between the signals within the categories. FFT Welch Method and Kohonen 's Self-Organizing Maps are then employed to determine the integrity level of the sources associated with the signals in category A. The overall results show 91.2% accuracy in classifying the signals into categories A and B. Additionally, the system was able to achieve 100% classification when distinguishing between the poor and (somewhat) good signal source integrity. The hybrid classification system proposed in this paper has direct application to real world problems where both signal isolation and the associated integrity of the signal source need to be determined.
机译：本文着重于一个独特的信号分类问题，该问题采用数字信号处理技术，首先将57个音频信号分为两个信号类型类别（A和B），然后将其分类为A类信号源的完整性。使用短时傅立叶变换和集中趋势分析来区分类别内的信号。然后，使用FFT Welch方法和Kohonen的自组织图确定与A类信号相关的信号源的完整性级别。总体结果表明，将信号分类为A和B类的准确性为91.2％。当区分差（和好）的信号源完整性时，该系统能够实现100％的分类。本文提出的混合分类系统可直接应用于现实世界中需要确定信号隔离度和信号源完整性的问题。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

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