Prior-free 3D human pose estimation in a video using limb-vectors

Estimating accurate 3D human poses from a monocular video is fundamental to various computer vision tasks. Existing methods exploit 2D-to-3D pose lifting, multiview images, and depth sensors to model spatio-temporal dependencies. However, depth ambiguities, occlusions, and larger temporal receptive...

Full description

Saved in:
Bibliographic Details
Main Authors: Anam Memon, Qasim Arain, Nasrullah Pirzada, Akram Shaikh, Adel Sulaiman, Mana Saleh Al Reshan, Hani Alshahrani, Asadullah Shaikh
Format: Article
Language:English
Published: Elsevier 2024-12-01
Series:ICT Express
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2405959524001188
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846129528631459840
author Anam Memon
Qasim Arain
Nasrullah Pirzada
Akram Shaikh
Adel Sulaiman
Mana Saleh Al Reshan
Hani Alshahrani
Asadullah Shaikh
author_facet Anam Memon
Qasim Arain
Nasrullah Pirzada
Akram Shaikh
Adel Sulaiman
Mana Saleh Al Reshan
Hani Alshahrani
Asadullah Shaikh
author_sort Anam Memon
collection DOAJ
description Estimating accurate 3D human poses from a monocular video is fundamental to various computer vision tasks. Existing methods exploit 2D-to-3D pose lifting, multiview images, and depth sensors to model spatio-temporal dependencies. However, depth ambiguities, occlusions, and larger temporal receptive fields pose challenges to these approaches. To address this, we propose a novel prior-free DCNN-based 3D human pose estimation method for monocular image sequences using limb vectors. Our method comprises two subnetworks: a limb direction estimator and a limb length estimator. The limb direction estimator utilizes a fully convolutional network to model limb direction vectors across a temporal window. We show that network complexity can be significantly reduced by utilizing dilated convolutional operations and a relatively smaller receptive field while maintaining estimation accuracy. Moreover, the limb length estimator captures stable limb length estimations from a reliable frame set. Our model has shown superior performance compared to existing methods on the Human3.6M and MPI-INF-3DHP datasets.
format Article
id doaj-art-4d4e9cc736774ddb8caad5d77d4c42ba
institution Kabale University
issn 2405-9595
language English
publishDate 2024-12-01
publisher Elsevier
record_format Article
series ICT Express
spelling doaj-art-4d4e9cc736774ddb8caad5d77d4c42ba2024-12-10T04:14:25ZengElsevierICT Express2405-95952024-12-0110612661272Prior-free 3D human pose estimation in a video using limb-vectorsAnam Memon0Qasim Arain1Nasrullah Pirzada2Akram Shaikh3Adel Sulaiman4Mana Saleh Al Reshan5Hani Alshahrani6Asadullah Shaikh7Department of Computer Systems Engineering, Mehran University of Engineering and Technology, Jamshoro, PakistanDepartment of Software Engineering, Mehran University of Engineering and Technology, Jamshoro, PakistanDepartment of Telecommunication Engineering, Mehran University of Engineering and Technology, Jamshoro, PakistanPASTIC National Centre, QAU Campus, Islamabad, PakistanDepartment of Computer Science, College of Computer Science and Information Systems, Najran University, Najran 61441, Saudi Arabia; Emerging Technologies Research Lab (ETRL), College of Computer Science and Information Systems, Najran University, Najran 61441, Saudi ArabiaDepartment of Information Systems, College of Computer Science and Information Systems, Najran University, Najran 61441, Saudi Arabia; Emerging Technologies Research Lab (ETRL), College of Computer Science and Information Systems, Najran University, Najran 61441, Saudi ArabiaDepartment of Computer Science, College of Computer Science and Information Systems, Najran University, Najran 61441, Saudi Arabia; Emerging Technologies Research Lab (ETRL), College of Computer Science and Information Systems, Najran University, Najran 61441, Saudi ArabiaDepartment of Information Systems, College of Computer Science and Information Systems, Najran University, Najran 61441, Saudi Arabia; Emerging Technologies Research Lab (ETRL), College of Computer Science and Information Systems, Najran University, Najran 61441, Saudi Arabia; Corresponding author at: Department of Information Systems, College of Computer Science and Information Systems, Najran University, Najran 61441, Saudi Arabia.Estimating accurate 3D human poses from a monocular video is fundamental to various computer vision tasks. Existing methods exploit 2D-to-3D pose lifting, multiview images, and depth sensors to model spatio-temporal dependencies. However, depth ambiguities, occlusions, and larger temporal receptive fields pose challenges to these approaches. To address this, we propose a novel prior-free DCNN-based 3D human pose estimation method for monocular image sequences using limb vectors. Our method comprises two subnetworks: a limb direction estimator and a limb length estimator. The limb direction estimator utilizes a fully convolutional network to model limb direction vectors across a temporal window. We show that network complexity can be significantly reduced by utilizing dilated convolutional operations and a relatively smaller receptive field while maintaining estimation accuracy. Moreover, the limb length estimator captures stable limb length estimations from a reliable frame set. Our model has shown superior performance compared to existing methods on the Human3.6M and MPI-INF-3DHP datasets.http://www.sciencedirect.com/science/article/pii/S24059595240011883D human pose estimationLimb vectorsPose correction
spellingShingle Anam Memon
Qasim Arain
Nasrullah Pirzada
Akram Shaikh
Adel Sulaiman
Mana Saleh Al Reshan
Hani Alshahrani
Asadullah Shaikh
Prior-free 3D human pose estimation in a video using limb-vectors
ICT Express
3D human pose estimation
Limb vectors
Pose correction
title Prior-free 3D human pose estimation in a video using limb-vectors
title_full Prior-free 3D human pose estimation in a video using limb-vectors
title_fullStr Prior-free 3D human pose estimation in a video using limb-vectors
title_full_unstemmed Prior-free 3D human pose estimation in a video using limb-vectors
title_short Prior-free 3D human pose estimation in a video using limb-vectors
title_sort prior free 3d human pose estimation in a video using limb vectors
topic 3D human pose estimation
Limb vectors
Pose correction
url http://www.sciencedirect.com/science/article/pii/S2405959524001188
work_keys_str_mv AT anammemon priorfree3dhumanposeestimationinavideousinglimbvectors
AT qasimarain priorfree3dhumanposeestimationinavideousinglimbvectors
AT nasrullahpirzada priorfree3dhumanposeestimationinavideousinglimbvectors
AT akramshaikh priorfree3dhumanposeestimationinavideousinglimbvectors
AT adelsulaiman priorfree3dhumanposeestimationinavideousinglimbvectors
AT manasalehalreshan priorfree3dhumanposeestimationinavideousinglimbvectors
AT hanialshahrani priorfree3dhumanposeestimationinavideousinglimbvectors
AT asadullahshaikh priorfree3dhumanposeestimationinavideousinglimbvectors