Cargando…
Three-Dimensional Human Pose Estimation from Sparse IMUs through Temporal Encoder and Regression Decoder
Three-dimensional (3D) pose estimation has been widely used in many three-dimensional human motion analysis applications, where inertia-based path estimation is gradually being adopted. Systems based on commercial inertial measurement units (IMUs) usually rely on dense and complex wearable sensors a...
Autores principales: | , , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
MDPI
2023
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10098669/ https://www.ncbi.nlm.nih.gov/pubmed/37050604 http://dx.doi.org/10.3390/s23073547 |
_version_ | 1785024865541029888 |
---|---|
author | Liao, Xianhua Dong, Jiayan Song, Kangkang Xiao, Jiangjian |
author_facet | Liao, Xianhua Dong, Jiayan Song, Kangkang Xiao, Jiangjian |
author_sort | Liao, Xianhua |
collection | PubMed |
description | Three-dimensional (3D) pose estimation has been widely used in many three-dimensional human motion analysis applications, where inertia-based path estimation is gradually being adopted. Systems based on commercial inertial measurement units (IMUs) usually rely on dense and complex wearable sensors and time-consuming calibration, causing intrusions to the subject and hindering free body movement. The sparse IMUs-based method has drawn research attention recently. Existing sparse IMUs-based three-dimensional pose estimation methods use neural networks to obtain human poses from temporal feature information. However, these methods still suffer from issues, such as body shaking, body tilt, and movement ambiguity. This paper presents an approach to improve three-dimensional human pose estimation by fusing temporal and spatial features. Based on a multistage encoder–decoder network, a temporal convolutional encoder and human kinematics regression decoder were designed. The final three-dimensional pose was predicted from the temporal feature information and human kinematic feature information. Extensive experiments were conducted on two benchmark datasets for three-dimensional human pose estimation. Compared to state-of-the-art methods, the mean per joint position error was decreased by 13.6% and 19.4% on the total capture and DIP-IMU datasets, respectively. The quantitative comparison demonstrates that the proposed temporal information and human kinematic topology can improve pose accuracy. |
format | Online Article Text |
id | pubmed-10098669 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2023 |
publisher | MDPI |
record_format | MEDLINE/PubMed |
spelling | pubmed-100986692023-04-14 Three-Dimensional Human Pose Estimation from Sparse IMUs through Temporal Encoder and Regression Decoder Liao, Xianhua Dong, Jiayan Song, Kangkang Xiao, Jiangjian Sensors (Basel) Article Three-dimensional (3D) pose estimation has been widely used in many three-dimensional human motion analysis applications, where inertia-based path estimation is gradually being adopted. Systems based on commercial inertial measurement units (IMUs) usually rely on dense and complex wearable sensors and time-consuming calibration, causing intrusions to the subject and hindering free body movement. The sparse IMUs-based method has drawn research attention recently. Existing sparse IMUs-based three-dimensional pose estimation methods use neural networks to obtain human poses from temporal feature information. However, these methods still suffer from issues, such as body shaking, body tilt, and movement ambiguity. This paper presents an approach to improve three-dimensional human pose estimation by fusing temporal and spatial features. Based on a multistage encoder–decoder network, a temporal convolutional encoder and human kinematics regression decoder were designed. The final three-dimensional pose was predicted from the temporal feature information and human kinematic feature information. Extensive experiments were conducted on two benchmark datasets for three-dimensional human pose estimation. Compared to state-of-the-art methods, the mean per joint position error was decreased by 13.6% and 19.4% on the total capture and DIP-IMU datasets, respectively. The quantitative comparison demonstrates that the proposed temporal information and human kinematic topology can improve pose accuracy. MDPI 2023-03-28 /pmc/articles/PMC10098669/ /pubmed/37050604 http://dx.doi.org/10.3390/s23073547 Text en © 2023 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). |
spellingShingle | Article Liao, Xianhua Dong, Jiayan Song, Kangkang Xiao, Jiangjian Three-Dimensional Human Pose Estimation from Sparse IMUs through Temporal Encoder and Regression Decoder |
title | Three-Dimensional Human Pose Estimation from Sparse IMUs through Temporal Encoder and Regression Decoder |
title_full | Three-Dimensional Human Pose Estimation from Sparse IMUs through Temporal Encoder and Regression Decoder |
title_fullStr | Three-Dimensional Human Pose Estimation from Sparse IMUs through Temporal Encoder and Regression Decoder |
title_full_unstemmed | Three-Dimensional Human Pose Estimation from Sparse IMUs through Temporal Encoder and Regression Decoder |
title_short | Three-Dimensional Human Pose Estimation from Sparse IMUs through Temporal Encoder and Regression Decoder |
title_sort | three-dimensional human pose estimation from sparse imus through temporal encoder and regression decoder |
topic | Article |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10098669/ https://www.ncbi.nlm.nih.gov/pubmed/37050604 http://dx.doi.org/10.3390/s23073547 |
work_keys_str_mv | AT liaoxianhua threedimensionalhumanposeestimationfromsparseimusthroughtemporalencoderandregressiondecoder AT dongjiayan threedimensionalhumanposeestimationfromsparseimusthroughtemporalencoderandregressiondecoder AT songkangkang threedimensionalhumanposeestimationfromsparseimusthroughtemporalencoderandregressiondecoder AT xiaojiangjian threedimensionalhumanposeestimationfromsparseimusthroughtemporalencoderandregressiondecoder |