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

Human body inverse dynamics solving method and device based on deep learning

An inverse dynamics and deep learning technology, applied in the field of computer graphics and computer animation, can solve the problems of unnatural human posture and lack of realism, and achieve the effect of natural posture and more posture.

Pending Publication Date: 2021-11-12
TSINGHUA UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] This application provides a method, device and storage medium for solving human body inverse dynamics based on deep learning, so as to at least solve the technical problems of unnatural human body posture and lack of realism in related technologies

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 body inverse dynamics solving method and device based on deep learning
  • Human body inverse dynamics solving method and device based on deep learning
  • Human body inverse dynamics solving method and device based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0032] figure 1 It is a schematic flow chart of the human body inverse dynamics solution method based on deep learning provided according to an embodiment of the present application, such as figure 1 As shown, can include:

[0033] Step S1. Obtain the coordinates of the target's human body end joints.

[0034] Wherein, the terminal joints in this application include joints corresponding to the hands, feet and head of the human body.

[0035] Step S2, according to the coordinates of the end joints of the target human body, the coordinates of the joints of the whole body are complemented by using the deep fully connected network.

[0036] Among them, the coordinates of the end joints of the human body are the coordinate difference between the corresponding end joints and the root node in the world coordinate system, so as to exclude the influence of the position of the root node.

[0037] And, the present application first randomly selects the global rotation of the root node...

Embodiment 2

[0049] Further, based on the deep learning-based human body inverse dynamics solution method provided in the above embodiments, the embodiment of the present application also provides a deep learning-based human body inverse dynamics solution device 200, figure 2 It is a schematic structural diagram of a human body inverse dynamics solving device based on deep learning provided according to an embodiment of the present application, such as figure 2 As shown, can include:

[0050] The acquiring module 201 is configured to acquire the coordinates of the end joints of the human body of the target.

[0051] The completion module 202 is configured to complete the coordinates of the joints of the whole body by using the deep fully connected network according to the coordinates of the end joints of the target.

[0052] The estimation module 203 is used for estimating the quaternion representation of the relative rotation of the whole body joints, that is, the posture of the human ...

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 provides a human body inverse dynamics solving method based on deep learning, and the method comprises the steps: firstly complementing the coordinates of the whole-body joints through a deep full-connection network for the obtained coordinates of the tail end joints of a target human body, and then estimating the posture of the human body through the deep full-connection network according to the coordinates of the whole-body joints; and judging whether the error between the position corresponding to the estimated coordinates of the tail end joint of the human body and the position corresponding to the coordinates of the tail end joint of the human body required to be reached in the input reaches the set precision or not, if so, rendering, otherwise, updating the posture by utilizing a cyclic coordinate descent algorithm, and re-judging until the rendering is successful. The global rotation of the root node is not limited, the global rotation can be estimated, and the estimated attitude is optimized through the cyclic coordinate descent algorithm, so that the finally obtained attitude is more natural and distortionless, and meanwhile, the obtained attitude can be rendered in real time to obtain a rendering result.

Description

technical field [0001] This application relates to the technical fields of computer graphics and computer animation, and in particular to a method, device and storage medium for solving human body inverse dynamics based on deep learning. Background technique [0002] Human body inverse dynamics is to obtain the most probable skeleton posture of the human body through the reverse movement of known human hands or feet. Since the same position of the hands, feet and head of the human body can have many different postures of the human body, the solution to inversely obtain the posture of the human body only through the positions of these joints is not unique. [0003] In related technologies, the posture of the human body calculated by the built-in inverse dynamics solver in the Unity3D game engine is unnatural and lacks a sense of reality. Contents of the invention [0004] The present application provides a human body inverse dynamics solution method, device and storage med...

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): G06T17/00G06T15/00G06N3/04G06N3/08
CPCG06T17/00G06T15/005G06N3/08G06N3/045
Inventor 徐枫伊昕宇
Owner TSINGHUA 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