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

Human skeleton behavior recognition method based on multi-stream fast and slow graph convolutional network

A convolutional network and recognition method technology, applied in character and pattern recognition, biological neural network models, instruments, etc., can solve the problem of high correlation of bone data, achieve enhanced information interaction, enhanced information extraction capabilities, and reduced number of channels Effect

Active Publication Date: 2020-10-30
NANJING UNIV OF POSTS & TELECOMM
View PDF2 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In the spatial dimension, the skeleton data can represent the current posture of the human body with a small number of joint points; but in the temporal dimension, the skeleton data still has a lot of redundant information, and the skeleton data of adjacent frames are highly correlated

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
  • Human skeleton behavior recognition method based on multi-stream fast and slow graph convolutional network
  • Human skeleton behavior recognition method based on multi-stream fast and slow graph convolutional network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0028] In order to illustrate the technical scheme of the present invention more clearly, the present invention will be further described below; Obviously, what is described below is only a part of the embodiment, for those of ordinary skill in the art, without paying creative work Under the premise, the technical solution of the present invention can also be applied to other similar scenarios according to these; in order to illustrate the technical solution of the present invention more clearly, the technical solution of the present invention will be further described in detail below in conjunction with the accompanying drawings:

[0029] As shown in the figure; a human skeleton behavior recognition method based on a multi-stream fast and slow graph convolutional network, including the following steps:

[0030] Step (1.1), creating a skeletal sequence behavior database of the human body, using a pose estimation algorithm to extract the skeletal joint points of each human body ...

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 human skeleton behavior recognition method based on a multi-stream fast and slow graph convolutional network. The invention relates to the technical field of image recognition, the thought of a fast and slow network is combined with a graph convolution network, the fast network can effectively extract the time information of a skeleton sequence, the slow network can effectively extract the space information of the skeleton sequence, and the lateral connection mode enhances the information interaction between two networks. By applying different attention mechanisms, the extraction and integration of spatial and temporal features are enhanced. Due to the fact that sampling is conducted and the number of channels is reduced, the calculated amount is greatly reduced.By introducing a multi-stream structure, the spatial information extraction capability is further enhanced, and the recognition rate and robustness of the behavior recognition method based on the bonejoint points are improved.

Description

technical field [0001] The invention relates to the technical field of image recognition, in particular to a human skeleton sequence behavior recognition method based on a multi-stream fast-slow graph convolutional network. Background technique [0002] Behavior recognition plays an important role in many applications such as intelligent video surveillance, automatic driving, human-computer interaction, and motion analysis. According to the input data type, action recognition can be roughly divided into two categories: RGB image sequences and bone sequences. For RGB image sequences, spatial appearance and temporal optical flow are usually used to model human behavior, however, human appearance in RGB image sequences is easily affected by factors such as illumination, viewing angle, and background. Skeleton sequence is a collection of human joint points in time and space. Compared with RGB image sequence, it has the advantages of small data volume and less susceptible to int...

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): G06K9/00G06K9/62G06N3/04
CPCG06V40/23G06N3/045G06F18/214
Inventor 孙宁冷令李晓飞
Owner NANJING UNIV OF POSTS & TELECOMM
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