Mitigating the Impact of Labeling Inaccuracies on 3D Human Body Reconstruction from Monocular Videos
Abstract This paper addresses the challenge of labeling inaccuracies in 3D human pose and shape reconstruction from monocular videos. Existing methods often rely on noisy pseudo ground truth, which introduces performance degradation such as jittering and drifting. To overcome these limitations, we p...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2025-07-01
|
| Series: | International Journal of Computational Intelligence Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s44196-025-00921-5 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849761262251016192 |
|---|---|
| author | Yupeng Hou Guangping Zeng |
| author_facet | Yupeng Hou Guangping Zeng |
| author_sort | Yupeng Hou |
| collection | DOAJ |
| description | Abstract This paper addresses the challenge of labeling inaccuracies in 3D human pose and shape reconstruction from monocular videos. Existing methods often rely on noisy pseudo ground truth, which introduces performance degradation such as jittering and drifting. To overcome these limitations, we propose a confidence-aware framework grounded in biomechanics. Our method adopts the SKEL model to provide anatomically constrained pose representations, reducing dependency on imprecise annotations. We further introduce a conditional normalizing flow to model per-parameter uncertainty conditioned on visual and motion features. Additionally, a novel evaluation metric, confidence-weighted procrustes aligned MPJPE, is proposed to incorporate confidence scores into performance assessment. Extensive experiments show that our approach outperforms existing methods on multiple datasets in both accuracy and motion smoothness. It demonstrates strong robustness against noisy annotations, and confidence estimates align closely with actual prediction errors. Ablation studies validate the contributions of both biomechanical modeling and confidence learning. Overall, our framework provides a unified and robust solution for 3D human reconstruction in real-world settings. |
| format | Article |
| id | doaj-art-91c8d2550b0a4dec943c238965e7daa3 |
| institution | DOAJ |
| issn | 1875-6883 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | Springer |
| record_format | Article |
| series | International Journal of Computational Intelligence Systems |
| spelling | doaj-art-91c8d2550b0a4dec943c238965e7daa32025-08-20T03:06:05ZengSpringerInternational Journal of Computational Intelligence Systems1875-68832025-07-0118111910.1007/s44196-025-00921-5Mitigating the Impact of Labeling Inaccuracies on 3D Human Body Reconstruction from Monocular VideosYupeng Hou0Guangping Zeng1School of Computer and Communication Engineering, University of Science and Technology BeijingSchool of Computer and Communication Engineering, University of Science and Technology BeijingAbstract This paper addresses the challenge of labeling inaccuracies in 3D human pose and shape reconstruction from monocular videos. Existing methods often rely on noisy pseudo ground truth, which introduces performance degradation such as jittering and drifting. To overcome these limitations, we propose a confidence-aware framework grounded in biomechanics. Our method adopts the SKEL model to provide anatomically constrained pose representations, reducing dependency on imprecise annotations. We further introduce a conditional normalizing flow to model per-parameter uncertainty conditioned on visual and motion features. Additionally, a novel evaluation metric, confidence-weighted procrustes aligned MPJPE, is proposed to incorporate confidence scores into performance assessment. Extensive experiments show that our approach outperforms existing methods on multiple datasets in both accuracy and motion smoothness. It demonstrates strong robustness against noisy annotations, and confidence estimates align closely with actual prediction errors. Ablation studies validate the contributions of both biomechanical modeling and confidence learning. Overall, our framework provides a unified and robust solution for 3D human reconstruction in real-world settings.https://doi.org/10.1007/s44196-025-00921-53D human body reconstruction3D human body modelConfidence modelingMonocular videos |
| spellingShingle | Yupeng Hou Guangping Zeng Mitigating the Impact of Labeling Inaccuracies on 3D Human Body Reconstruction from Monocular Videos International Journal of Computational Intelligence Systems 3D human body reconstruction 3D human body model Confidence modeling Monocular videos |
| title | Mitigating the Impact of Labeling Inaccuracies on 3D Human Body Reconstruction from Monocular Videos |
| title_full | Mitigating the Impact of Labeling Inaccuracies on 3D Human Body Reconstruction from Monocular Videos |
| title_fullStr | Mitigating the Impact of Labeling Inaccuracies on 3D Human Body Reconstruction from Monocular Videos |
| title_full_unstemmed | Mitigating the Impact of Labeling Inaccuracies on 3D Human Body Reconstruction from Monocular Videos |
| title_short | Mitigating the Impact of Labeling Inaccuracies on 3D Human Body Reconstruction from Monocular Videos |
| title_sort | mitigating the impact of labeling inaccuracies on 3d human body reconstruction from monocular videos |
| topic | 3D human body reconstruction 3D human body model Confidence modeling Monocular videos |
| url | https://doi.org/10.1007/s44196-025-00921-5 |
| work_keys_str_mv | AT yupenghou mitigatingtheimpactoflabelinginaccuracieson3dhumanbodyreconstructionfrommonocularvideos AT guangpingzeng mitigatingtheimpactoflabelinginaccuracieson3dhumanbodyreconstructionfrommonocularvideos |