北京航空航天大学学报 ›› 2020, Vol. 46 ›› Issue (9): 1677-1681.doi: 10.13700/j.bh.1001-5965.2020.0046

• 论文 • 上一篇    下一篇

基于迁移学习的暴恐图像自动识别

陈猛夫   

  1. 北京航空航天大学 公共管理学院, 北京 100083
  • 收稿日期:2020-02-22 发布日期:2020-09-22
  • 通讯作者: 陈猛夫 E-mail:dhzhou2084@163.com
  • 作者简介:陈猛夫 男,硕士,高级工程师。主要研究方向:数据挖掘、信息融合。

Automatic recognition for terrorism related image based on transfer learning

CHEN Mengfu   

  1. School of Public Administration, Beihang University, Beijing 100083, China
  • Received:2020-02-22 Published:2020-09-22

摘要: 利用人工智能和深度学习技术自动化地分析互联网海量图片,快速、准确地识别有害的暴恐图像并及时处置是反恐工作的重要手段之一。研究了利用深度学习和迁移学习技术对暴恐图像进行分类识别。首先,定义了暴恐图像的主要概念特征,并针对性地构建数据集;其次,针对暴恐图像正样本较少的问题,设计深度神经网络模型和迁移学习方式;最后,基于构建的训练数据集进行模型训练和测试。结果显示:所提方法可以快速、准确地对互联网图片进行分类识别,平均分类准确率达到96.7%,从而有效降低人工检测的劳动强度,为反恐预警工作提供决策支持。

关键词: 暴恐图像, 深度学习, 图像识别, 卷积神经网络, 迁移学习

Abstract: Using AI and deep learning technology to automatically analyze massive Internet pictures, quickly and accurately identifying harmful images related to terrorism and dealing with them in time is one of the important means for anti-terrorism work. This paper studies how to use deep learning and transfer learning technology to classify and recognize the images related to terrorism. First, we define the main concept features of the image related to terrorism and collect the relevant positive samples to construct dataset. Second, we design suitable deep neural network model and transfer learning method for the problem of less positive samples of the image related to terrorism. Finally, using the constructed training dataset to fine-tune the final model. The results show that, based on the proposed method in this paper, we can classify and recognize the Internet pictures which have terrorism content quickly and accurately with average classification accuracy rate of 96.7%, and thus the labor intensity of manual monitoring will reduce effectively, which can provide support for decision-making in the work of anti-terrorism early warning.

Key words: terrorism related image, deep learning, image recognition, convolutional neural network, transfer learning

中图分类号: 


版权所有 © 《北京航空航天大学学报》编辑部
通讯地址:北京市海淀区学院路37号 北京航空航天大学学报编辑部 邮编:100191 E-mail:jbuaa@buaa.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发