MOTION RECOGNITION AND GESTURE PREDICTION METHOD AND DEVICE

This application provides an action recognition and pose estimation method and apparatus. The action recognition method includes: determining a to-be-processed video, where the to-be-processed video is a video that includes images of a human body; determining a to-be-processed image based on the to-...

Full description

Saved in:
Bibliographic Details
Main Authors WANG, Yali, JIANG, Lihui, QIAO, Yu, DU, Wenbin, LIU, Jianzhuang
Format Patent
LanguageEnglish
French
German
Published 05.08.2020
Subjects
CALCULATING
COMPUTING
COUNTING
HANDLING RECORD CARRIERS
PHYSICS
PRESENTATION OF DATA
RECOGNITION OF DATA
RECORD CARRIERS
Online AccessGet full text

Cover

More Information
Summary:This application provides an action recognition and pose estimation method and apparatus. The action recognition method includes: determining a to-be-processed video, where the to-be-processed video is a video that includes images of a human body; determining a to-be-processed image based on the to-be-processed video; performing convolution on the to-be-processed image to obtain a plurality of eigenvectors, where the plurality of eigenvectors are respectively used to denote features of different locations in the to-be-processed image; determining a weight coefficient set of each human joint of the human body based on the plurality of eigenvectors; weighting the plurality of eigenvectors based on the weight coefficient set of each human joint to obtain an action feature of each human joint; determining an action feature of the human body based on the action feature of each human joint; and determining an action type of the human body based on the action feature of the human body. This application can improve action recognition accuracy.
Bibliography:Application Number: EP20180866516