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

Real-time key point detection method based on lightweight neural network

A neural network and detection method technology, applied in the field of key point detection of computer vision, can solve problems such as running speed constraints, and achieve satisfactory detection accuracy

Active Publication Date: 2019-10-25
ZHEJIANG UNIV
View PDF8 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0002] Key point detection technology has extremely important applications in the fields of human action recognition, pedestrian re-identification and related fields, but whether it is in security or consumer fields such as somatosensory games, there are strict requirements on the running time of the system. The current mainstream Although the key point detection algorithm of the present invention has obtained a huge improvement in the detection accuracy, it has been severely restricted in the running speed, so the present invention mainly realizes that the The detection accuracy of the actual application, in order to improve the running speed under the premise of ensuring the detection accuracy

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
  • Real-time key point detection method based on lightweight neural network
  • Real-time key point detection method based on lightweight neural network
  • Real-time key point detection method based on lightweight neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0021] The method of the present invention will be further described below in conjunction with the accompanying drawings and specific examples.

[0022] The following examples are only used to further illustrate the method of the present invention, but not specifically limit the present invention. figure 1 A structural representation (c=4) of a concrete lightweight neural network built for the present invention, based on this lightweight neural network, the real-time key point detection method is carried out, and the steps are as follows:

[0023] First, the human body in the image is intercepted by a deep learning-based human detection technology (such as faster rcnn, etc.), and then the result obtained by the interception is sent to the constructed lightweight neural network. The lightweight neural network includes several bottlenecks Module (bottleneck) A and bottleneck module B, where the step size of the 3×3 convolutional layer in bottleneck module A is 1, and the step si...

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 real-time key point detection method based on a lightweight neural network. The method comprises the following steps: firstly, intercepting figures in a picture by using a faster rcnn or similar human body detection algorithm, and then sending the figures into a designed lightweight network, and outputting a predicated coordinate. Different from a common method, the method has the advantages that the designed network is lightweight and telescopic, satisfactory detection precision can be achieved with small parameter quantity and calculation quantity, and meanwhile, end-to-end prediction is completed by the system and 2D and 3D tasks can be completed by using integral-based heat map prediction.

Description

technical field [0001] The invention belongs to the field of key point detection of computer vision, and in particular relates to a real-time key point detection method of a lightweight neural network. Background technique [0002] Key point detection technology has extremely important applications in the fields of human action recognition, pedestrian re-identification and related fields, but whether it is in security or consumer fields such as somatosensory games, there are strict requirements on the running time of the system. The current mainstream Although the key point detection algorithm of the present invention has obtained a huge improvement in the detection accuracy, it has been severely restricted in the running speed, so the present invention mainly realizes that the The detection accuracy of the actual application is used to improve the running speed while ensuring the detection accuracy. Contents of the invention [0003] The object of the present invention i...

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
IPC IPC(8): G06K9/00G06K9/46G06K9/62G06N3/04G06N3/08G06T3/40
CPCG06N3/08G06T3/4023G06T2207/10004G06T2207/30196G06V40/10G06V10/462G06N3/045G06F18/253
Inventor 王雷黄科杰
Owner ZHEJIANG UNIV
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