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

HRNet human body posture recognition method based on attention mechanism optimization

A technology of human body posture and recognition method, applied in the field of human body posture recognition, can solve the problems of low accuracy and high computing cost, and achieve the effect of improving performance, ensuring recognition accuracy and saving computing cost.

Pending Publication Date: 2022-04-29
JIANGNAN UNIV
View PDF0 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problems of high computational cost and low precision of current human body posture recognition methods, the present invention provides an HRNet human body posture recognition method based on attention mechanism optimization

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
  • HRNet human body posture recognition method based on attention mechanism optimization
  • HRNet human body posture recognition method based on attention mechanism optimization
  • HRNet human body posture recognition method based on attention mechanism optimization

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0103] The present embodiment provides a human body gesture recognition system based on an optimized HRNet network, the recognition system comprising: a video stream acquisition module, an optimized HRNet network module, and a classification result output module; the video stream acquisition module, an optimized HRNet network module, The classification result output modules are connected sequentially;

[0104] The optimized HRNet network module includes: a basic HRNet module, an expansion convolution module, and an attention mechanism module; the expansion convolution module is located between different resolution subnetworks in the basic HRNet module, and is used for The up-sampling or down-sampling process between sub-networks increases the receptive field; the attention mechanism module is used for weighted fusion of feature maps of different resolutions.

Embodiment 2

[0106] This embodiment provides a human body posture recognition method based on an optimized HRNet network, the method is realized based on the human body posture recognition system recorded in Embodiment 1, including:

[0107] Step 1: Get the video stream;

[0108] Step 2: The optimized HRNet network module obtains the image of human body pose to be recognized in the video stream, and inputs the image to the original resolution channel to obtain a high-resolution subnet feature map, which is the output feature map of the first stage;

[0109] Step 3: Down-sampling the high-resolution feature map to obtain a low-resolution subnetwork feature map with a resolution of 1 / 2 times the original, and use dilated convolution to increase the receptive field during the sampling process;

[0110] Step 4: Introduce the attention mechanism to perform cross-resolution feature fusion on the feature maps of different resolution subnetworks, perform global average pooling on the feature maps ...

Embodiment 3

[0114] This embodiment provides a human body posture recognition method based on an optimized HRNet network, the method is realized based on the human body posture recognition system recorded in Embodiment 1, including:

[0115] Step 1: Obtain the video stream. The method for obtaining the video stream in this embodiment is video reading;

[0116] Step 2: The optimized HRNet network module obtains the image of human body pose to be recognized in the video stream, and inputs the image to the original resolution channel to obtain a high-resolution subnet feature map, which is the output feature map of the first stage;

[0117] Step 3: Down-sampling the high-resolution feature map to obtain a low-resolution subnetwork feature map with a resolution of 1 / 2 times the original, and use dilated convolution to increase the receptive field during the sampling process;

[0118] In this embodiment, the calculation method of the dilated convolution input and output feature maps is:

[011...

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 an HRNet human body posture recognition method based on attention mechanism optimization, and belongs to the field of human body posture recognition and deep learning processing. According to the method, firstly, expansion convolution is added in the cross-channel fusion process of different-resolution feature maps, a receptive field can be increased without changing the size of a low-resolution feature map under the condition that extra parameters and calculation amount are not generated, and it is ensured that important information is not ignored when a decision is made; secondly, providing a new feature fusion strategy, performing weighted fusion on feature maps with different resolutions by introducing a channel attention mechanism, and adaptively recalibrating the other direction of feature mapping to enhance meaningful features, inhibit weak features, accelerate convergence speed and optimize human body posture recognition performance; and the detection precision is further improved.

Description

technical field [0001] The invention relates to an HRNet human body posture recognition method optimized based on an attention mechanism, and belongs to the fields of human body posture recognition and deep learning processing. Background technique [0002] Human pose estimation is an important research topic in computer vision. At present, human pose can be inferred through the pose detection method Part Affinity Field (PAF, Part Affinity Field) and the combination of key points and human graphic structure. The algorithm first decomposes the human body structure into many nodes, uses the human body structure model to model the relationship between the nodes, and finally divides each node and connects them into a whole to form a complete human body posture. However, in the case of complex backgrounds and highly flexible human postures, the accuracy and efficiency of the structural model in the figure below for similar actions will drop sharply, making it difficult to reach t...

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): G06V40/20G06V20/40G06N3/04G06N3/08
CPCG06N3/08G06N3/045
Inventor 杨金龙冯雨刘佳张媛刘建军
Owner JIANGNAN UNIV
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com