Unlock instant, AI-driven research and patent intelligence for your innovation.

Method, system and device for identifying illegal images

An image and image feature technology, applied in the field of systems and equipment, illegal image recognition methods, can solve the problems of poor robustness, unable to recognize images, not considering high-level features, etc., to achieve the effect of improving performance and robustness

Active Publication Date: 2022-05-20
人民中科(北京)智能技术有限公司
View PDF18 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] This paper is used to solve the problem that the existing illegal image recognition method does not consider high-level features, and cannot identify images with text violations, which has low recognition accuracy and poor robustness.

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
  • Method, system and device for identifying illegal images
  • Method, system and device for identifying illegal images
  • Method, system and device for identifying illegal images

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0127] The following will clearly and completely describe the technical solutions in the embodiments herein in conjunction with the accompanying drawings in the embodiments herein. Obviously, the described embodiments are only some of the embodiments herein, not all of them. Based on the embodiments herein, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts fall within the scope of protection herein.

[0128] It should be noted that the terms "first" and "second" in the description and claims herein and the above drawings are used to distinguish similar objects, but not necessarily used to describe a specific order or sequence. It is to be understood that the data so used are interchangeable under appropriate circumstances such that the embodiments herein described herein can be practiced in sequences other than those illustrated or described herein. Furthermore, the terms "comprising" and "having", as well as any variations ...

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

This article relates to the field of artificial intelligence and provides a method, system and equipment for illegal image recognition. The method includes: analyzing the image to be recognized to obtain text information and image information; inputting the text information into the text feature extraction model to obtain the text feature vector group ; Input the image information into the image feature extraction model to obtain the image feature vector group; input the text feature vector group and the image feature vector group into the feature fusion model to obtain the fusion feature vector group, and the feature fusion model is used to utilize cross-attention The mechanism makes the text feature vector add attention to the image feature vector; the fusion feature vector group is input into the classification recognition model, and the violation recognition result is obtained. In this paper, the cross-attention mechanism is used to fuse two modal features, which can improve the performance and robustness of illegal image recognition.

Description

technical field [0001] This article relates to the field of artificial intelligence technology, and in particular to a method, system and device for identifying illegal images. Background technique [0002] There are a large number of illegal images on the Internet. The widespread dissemination of these images not only seriously endangers the physical and mental health of young people, but also increases social instability. Image recognition is one of the most basic research problems in the field of computer vision. In the prior art, the following methods are mainly used to identify illegal images: [0003] (1) Based on the method of manual feature extraction, this method is mainly aimed at the extraction of the underlying features of the image (such as RGB, HOG, SIFT, SURF, SILTP, etc.), the processing is relatively simple, and no learning and training are required; [0004] (2) The deep model based on the convolutional neural network recognizes the image, but the existing...

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): G06V10/80G06V30/148G06V10/774G06V10/764G06F40/289
CPCG06F40/289G06F18/24G06F18/253G06F18/214
Inventor 阮晓峰王坚李兵余昊楠胡卫明
Owner 人民中科(北京)智能技术有限公司