Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Image human motion recognition method based on hierarchical information transfer

A technology for human action recognition and information transmission, which is applied in the field of computer vision and can solve problems such as lack of analysis and modeling

Active Publication Date: 2019-01-15
广州中科智巡科技有限公司
View PDF8 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Obviously, Zhao's method lacks analysis and modeling in this area

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
  • Image human motion recognition method based on hierarchical information transfer
  • Image human motion recognition method based on hierarchical information transfer
  • Image human motion recognition method based on hierarchical information transfer

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0038] The present invention is based on the picture human action recognition method of hierarchical information transmission, mainly comprises the following steps:

[0039] (1) Divide the human body into a hierarchical structure, which is composed of local body regions with finer granularity from top to bottom, that is, recursively decompose the human body into smaller body parts;

[0040] (2) Build a hierarchical propagation network, recursively transfer and integrate the information of the hierarchical structure in step (1), so as to obtain the final action descriptor;

[0041] (3) Combine the action descriptor obtained in step (2) with additional full-image information, input the final fully connected layer for classification, use the sigmoid function to calculate the probability distribution of the confidence, and use binary cross-entropy to calculate the classification loss.

[0042] The technical solution of the present invention is verified on the public HICO (Human I...

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 picture human body action recognition method based on hierarchical information transmission, which comprises the following steps: S1, the human body is divided into a hierarchical structure, the hierarchical structure is composed of local body regions with finer and finer granularity from top to bottom, that is, the human body is recursively decomposed into smaller body parts; 2, constructing a hierarchical propagation network, transmitting and integrating the hierarchical structure information in the step S1 recursively, thereby obtaining a final action descriptor; S3, combining the action descriptor obtained in the step S2 with the extra full graph information, inputting the final full connection layer for classification, using the sigmoid function to calculatethe probability distribution of the confidence degree, and using the binary cross entropy to calculate the classification loss. The invention defines an abstract human body segmentation framework andsegmentation rules, which makes the selection of the human body segmentation scheme more flexible and reduces singularity or irrationality.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a picture human action recognition method based on hierarchical information transmission. Background technique [0002] Image action recognition is a basic and important research in the field of computer vision. It requires the model to recognize the actions performed by the human body in the image and give the category of the action. Image action recognition has many applications, such as image caption, collective activity recognition, human-object interaction and so on. Image action recognition is also the basis of video action analysis. By treating a video frame as a single image, the problem can be transformed into image action recognition, which can be further modeled in the temporal dimension. [0003] The existing work can be mainly divided into three categories: image action recognition based on human body pose, image action recognition based on hiera...

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): G06K9/00G06N3/04
CPCG06V40/20G06N3/045
Inventor 胡建芳朱海昇谢佳锋郑伟诗
Owner 广州中科智巡科技有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com