A matching detection method of network news pictures based on deep learning

A deep learning and detection method technology, applied in the field of information detection, can solve problems such as image-text mismatch, achieve the effect of improving accuracy and saving manpower and material resources

Active Publication Date: 2020-01-03
INNER MONGOLIA UNIV OF TECH
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the shortcomings of the above-mentioned prior art, the purpose of the present invention is to provide a method for detecting the matching of network news pictures based on deep learning, which generates descriptions for news pictures through deep learning technology, and then compares them with news text content. To judge whether the illustrations are consistent with the content of the news text, so as to solve the mismatch between the pictures and the text in the false information, remove the false information, and effectively purify the network environment

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A matching detection method of network news pictures based on deep learning
  • A matching detection method of network news pictures based on deep learning
  • A matching detection method of network news pictures based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] The implementation of the present invention will be described in detail below in conjunction with the drawings and examples.

[0041] refer to figure 1 , the simple workflow of the present invention is as follows;

[0042] 1. Extract the pictures and news text content of the news website, and form a standardized document for the news content: that is, a standardized single sentence.

[0043] 2. Set the network structure and model parameters for image description generation.

[0044] (1) Convolutional network structure settings:

[0045] Using the network structure of VGG-16, it contains 13 layers of convolutional layers with a convolution kernel of 3×3 and 4 layers of maximum pooling layers with a pooling kernel of 2×2. For a picture with a size of 3×W’×H’, after passing through the convolutional network, the output result is a feature matrix of C×W’×H’.

[0046] (2) Full convolution positioning layer settings:

[0047] 1) Input and output

[0048] Input: feature ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides a deep learning-based network news illustration matching detection method. The method comprises the following steps of generating a plurality of descriptions for a news illustration based on deep learning; performing comparison scoring on text content of the generated descriptions for the news illustration and text content of the news. For the part of generating the news illustration descriptions, the news illustration characteristics are extracted by adopting a convolutional neural network, and then the related descriptions of the news illustration are generated by utilizing a natural language model. For the scoring system part, due to the difference in content length and expression modes between the generated picture description and the news text, a solution is provided, and a scoring system is formed by the solution and an improved BLEU algorithm. The scoring system compares and scores the generated picture description and the news text content, and judges whether the picture is consistent with the news content or not through scoring, so that the false information with inconsistent images and texts can be found more quickly and more accurately, the manualauditing time is shortened, the manpower and material resources are saved, and the network environment is purified.

Description

technical field [0001] The invention belongs to the technical field of information detection, relates to detection of false information, and in particular to a method for detecting matching of network news pictures based on deep learning. Background technique [0002] The deep integration of digital technology represented by the Internet and various fields of today's society, and the large-scale popularization of mobile devices represented by smart phones have laid a solid foundation for the birth and development of mobile news clients. In order to attract readers, a large number of news websites and mobile phone clients have edited a large number of junk news, such as headlines, graphic mismatch, etc., resulting in the proliferation of false information and misleading the masses. [0003] The mismatch of graphics and text belongs to the category of false information. Currently, a manual reporting and review system is used to reduce the share of articles that do not match gr...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/958G06F16/583
Inventor 云静尚俊峰刘利民许志伟
Owner INNER MONGOLIA UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products