A ddnn based on cloud-edge collaborative computing and its construction method and application

An edge and edge-side technology, applied to DDNN based on cloud-edge collaborative computing and its construction field, can solve the problem of low overall performance of DDNN, achieve the effect of improving accuracy, improving classification accuracy, and increasing effectiveness

Active Publication Date: 2022-02-15
HUAZHONG UNIV OF SCI & TECH
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The present invention provides a DDNN based on cloud-edge collaborative computing and its construction method and application, which are used to solve the problems caused by the large amount of parameters and cloud-edge communication in the existing deep neural network model of "cloud-edge" collaborative computing. The technical problem of low overall performance of DDNN

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 ddnn based on cloud-edge collaborative computing and its construction method and application
  • A ddnn based on cloud-edge collaborative computing and its construction method and application
  • A ddnn based on cloud-edge collaborative computing and its construction method and application

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0060] A DDNN based on cloud-edge collaborative computing is applied to object classification under multi-view images of the Internet of Things, including edge side and cloud.

[0061] Among them, the edge side is used to extract the features of each viewing angle image, and the feature bag model is used to measure the similarity of the extracted feature images and obtain the histogram vector of the viewing angle; the histogram vector of each viewing angle Based on the fused histogram vector, the target classification on the edge side is obtained. If the target classification does not meet the accuracy requirements, multiple feature images of each view are transmitted to the cloud; the cloud is used for all Feature weighted fusion is performed between multiple feature images, and convolution and classification operations are performed on the fused feature images to obtain the target classification in the cloud.

[0062] This embodiment introduces a bag-of-features model on the...

Embodiment 2

[0089] A method 100 for constructing a DDNN based on cloud-edge collaborative computing as described in Embodiment 1 above, including:

[0090] Step 110, establishing the edge side regression model and the cloud regression model of DDNN respectively;

[0091] Step 120, based on the edge side regression model and the cloud regression model, according to the actual needs of classification accuracy, communication volume and communication time, determine the edge side parameters and cloud parameters, and obtain DDNN.

[0092] This embodiment provides a method for constructing DDNN. Specifically, the regression model of DDNN on the edge side and the regression model of cloud DDNN are introduced first. The actual limit of the amount and the actual limit of the communication time are beneficial to the regression model, obtaining the parameters of the edge side and the cloud, and then quickly and efficiently constructing the DDNN that is actually required. This construction method ca...

Embodiment 3

[0108] A target classification method under multi-view images of the Internet of Things, comprising:

[0109] Collect multi-view images of the target;

[0110] Based on the multi-view image, use any DDNN based on cloud-edge collaborative computing as described in the first embodiment above to obtain the target classification; or determine the edge side and the cloud based on the DDNN construction method described in the second embodiment above to obtain target classification.

[0111] Using the above-mentioned DDNN can reduce the resource occupation of the edge side and improve the classification accuracy. The relevant technical solutions are the same as those in Embodiment 1, and will not be repeated here.

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 DDNN based on cloud-edge collaborative computing and its construction method and application, which are applied to object classification under multi-view images of the Internet of Things. DDNN includes: an edge side, which adopts a feature bag model to extract multiple features The similarity of the image is measured and the histogram vector of each view is obtained statistically; the feature weighted fusion is performed on the histogram vector of each view, and the classification of the edge side target is obtained based on the fused histogram vector. If the classification accuracy is not enough, the extracted The feature images of each viewing angle are transmitted to the cloud; the cloud is used to carry out feature weighted fusion of all feature images, perform convolution and classification operations on the fused feature images, and obtain the target classification in the cloud. The present invention introduces a feature bag model on the edge side to reduce the amount of parameters; in addition, multi-view weighted feature fusion reduces multi-view redundant features and increases the effectiveness of feature expression capabilities. The invention reduces the amount of parameters of DDNN and the traffic of cloud-edge communication and improves the overall performance of DDNN.

Description

technical field [0001] The invention belongs to the technical field of Internet of Things big data information extraction, and more specifically relates to a DDNN based on cloud-edge collaborative computing and its construction method and application. Background technique [0002] Internet of Things (IoT) devices are deployed in complex environments, for example, sensors are geographically distributed and can acquire multi-view data of a target object. Therefore, the multi-camera sensor network can obtain information from different perspectives of the same object, and making full use of this data reflects the diversity and value of IoT big data. [0003] Deep learning is one of the most promising approaches to extract precise information from raw IoT big data. Distributed Deep Neural Network (DDNN) has a multi-layer structure, and the amount of calculation parameters increases exponentially with the increase of the number of network layers. The most prominent performance is...

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): G06V10/764G06V10/80G06V10/74G06V10/82G06K9/62G06V10/766G06N3/04
CPCG06V10/50G06N3/045G06F18/22G06F18/241G06F18/253
Inventor 肖江文邹颖王燕舞
Owner HUAZHONG UNIV OF SCI & 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