A Hierarchical Children’s Dance Movement Pose Estimation Method Based on Sequence Multiscale Feature Fusion Representation

Aiming at the problem that traditional human motion pose estimation methods cannot accurately capture and estimate the movement changes of children dancers, a hierarchical dance pose estimation method for children based on sequence multiscale feature fusion representation is proposed. By comparing t...

Full description

Saved in:
Bibliographic Details
Published inAdvances in Multimedia Vol. 2022; pp. 1 - 10
Main Authors Qin, Yanan, Huang, Tao, Tang, Guanzhen
Format Journal Article
LanguageEnglish
Published New York Hindawi 24.08.2022
John Wiley & Sons, Inc
Hindawi Limited
Subjects
Algorithms
Analysis
Children's furniture
Dance
Dancers & choreographers
Dancing
Datasets
Feature extraction
Human body
Human motion
Methods
Nightclubs
Pose estimation
Representations
China
Online AccessGet full text

Cover

More Information
Summary:Aiming at the problem that traditional human motion pose estimation methods cannot accurately capture and estimate the movement changes of children dancers, a hierarchical dance pose estimation method for children based on sequence multiscale feature fusion representation is proposed. By comparing the pose feature extraction algorithm with the actual recognition effect, the recognition rates of the dancer’s upper body, infiltration, and whole body have increased by 14.2, 10.6, and 12.6, respectively. The experimental results show that the proposed pose estimation algorithm achieves good pose estimation results on both the standard human pose estimation dataset and the self-built dance dataset.
ISSN:1687-5680
1687-5699
DOI:10.1155/2022/2445210