Defense method for generating general inverse disturbance based on generative adversarial

A generative and model-generating technology, applied in neural learning methods, biological neural network models, neural architectures, etc., can solve problems such as defense methods that are difficult to effectively resist

Inactive Publication Date: 2020-04-07
ZHEJIANG UNIV OF TECH
View PDF0 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In the face of general disturbances, the above defense methods are difficult to effectively resist

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
  • Defense method for generating general inverse disturbance based on generative adversarial
  • Defense method for generating general inverse disturbance based on generative adversarial
  • Defense method for generating general inverse disturbance based on generative adversarial

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0047] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be pointed out that the embodiments described below are intended to facilitate the understanding of the present invention and do not have any limiting effect on it.

[0048] Such as figure 1 As shown, a general inverse perturbation defense method based on generative confrontation includes the following steps:

[0049] Step 1, build a suitable generative confrontation network structure GAN.

[0050] There are two models in GAN: generative model G and discriminant model D. The generative model G learns the characteristic distribution of general inverse disturbance through a large number of parameters of the neural network and captures the data distribution of anti-interference samples. Discriminant model D is the identification and detection model. To estimate whether a sample is a pure sample, the recognition ability of the discriminant model D ...

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 discloses a defense method for generating general inverse disturbance based on generative adversarial, and the method comprises the steps: (1) building a generative adversarial network which comprises a generation model and a discrimination model; (2) obtaining a relatively comprehensive adversarial sample by utilizing various attack methods; (3) pre-training the generation model andthe discrimination model in the generative adversarial network by adopting normal samples; (4) establishing a loss function of the generation model, and training the generation model; (5) establishing a loss function of the discrimination model, and training the discrimination model; (6) repeating the steps (4) and (5) until the number of iterations reaches a preset upper limit value or the lossfunctions of the two models reach a preset threshold value; and (7) performing performance index detection and application on the trained generation model. According to the method, a proper generativeadversarial network is built, feature extraction of general disturbance distribution is completed, and proper general inverse disturbance is generated, so that the robustness of the model is improved.

Description

Technical field [0001] The invention belongs to the field of deep learning security technology, and in particular relates to a defense method for generating general inverse disturbance based on generative confrontation. Background technique [0002] Deep learning can obtain more accurate classification results than general algorithms by learning and calculating the potential connections of large amounts of data, and has powerful feature learning capabilities and feature expression capabilities. Therefore, deep learning technology is widely used in the field of artificial intelligence, including autonomous driving technology, augmented reality technology, computer vision, biomedical diagnosis, natural language processing technology, etc. Deep learning uses a neural network with huge parameters to perform feature extraction to complete the fitting of a large amount of data distribution, thereby showing good image processing capabilities. [0003] At present, deep learning technology...

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 Applications(China)
IPC IPC(8): G06N3/04G06N3/08
CPCG06N3/088G06N3/045
Inventor 陈晋音朱伟鹏吴长安
Owner ZHEJIANG 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

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap