Convolution-Based Design for Real-Time Pose Recognition and Character Animation Generation
Human pose recognition and its generation are an important animation design key point. To this end, this paper designs new neural network structures for 2D and 3D pose extraction tasks and corresponding GPU-oriented acceleration schemes. The scheme first takes an image as input, extracts the human p...
Main Authors: | Lee, J. (Author), Wang, D. (Author) |
---|---|
Format: | Article |
Language: | English |
Published: |
Hindawi Limited
2022
|
Subjects: | |
Online Access: | View Fulltext in Publisher |
Similar Items
-
The Progress of Human Pose Estimation: A Survey and Taxonomy of Models Applied in 2D Human Pose Estimation
by: Tewodros Legesse Munea, et al.
Published: (2020-01-01) -
Real-Time Human Pose Estimation and Gesture Recognition from Depth Images Using Superpixels and SVM Classifier
by: Hanguen Kim, et al.
Published: (2015-05-01) -
Kan man identifiera de karaktäristiska dragen hos stop-motion och applicera dem inom 3D-animation?
by: Dahlin Jansson, Emelie
Published: (2013) -
Simple Baseline for Vehicle Pose Estimation: Experimental Validation
by: Hector Corrales Sanchez, et al.
Published: (2020-01-01) -
Crater Detection and Recognition Method for Pose Estimation
by: Zihao Chen, et al.
Published: (2021-09-01)