HADD: A hierarchy-aware disentangled diffusion framework with spatio-temporal denoising for 3D human pose estimation
6 affiliations
- 1Department of Physical Education, Fujian Agriculture and Forestry University, Fuzhou, China
- 2College of Intelligent Science and Technology, Inner Mongolia University of Technology, Huhhot, China
- 3Department of Physical Education, Fujian University of Technology, Fuzhou, China
- 4School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore, Singapore
- 5School of Software Engineering, Sichuan Polytechnic University, Deyang, China
- 6Department of Mechanical Engineering, Hebei Petroleum University of Technology, Chengde, China
Abstract
Monocular 3D Human Pose Estimation (3D HPE) is a core task in computer vision, which provides critical technical support for applications including VR/AR, immersive human-computer interaction, digital human animation, as well as sports motion capture, athletic technique diagnosis, physical fitness assessment and sports rehabilitation training. However, existing methods still face three key limitations: diffusion-based methods lack explicit human anatomical prior constraints in the diffusion process; disentangled-based methods suffer from severe hierarchical error accumulation along the skeletal kinematic tree; and most mainstream methods fail to fully mine the fine-grained hierarchical spatio-temporal correlations of the human skeleton, leading to increased estimation error for high-hierarchy joints. To address these challenges, we propose a novel Hierarchy-Aware Disentangled Diffusion framework with Spatio-Temporal Denoising (HADD) for monocular 3D HPE, which deeply integrates disentanglement strategy with the diffusion model and embeds skeletal hierarchical information into the full diffusion pipeline. In the forward diffusion process, HADD disentangles 3D pose into bone length and bone direction features, and injects Gaussian noise into the two features separately. In the reverse denoising process, we design a Hierarchical Spatial and Temporal Denoising (HSTD) module to accurately model the hierarchical spatio-temporal relationships of the human skeleton and alleviate error amplification of high-hierarchy joints. Meanwhile, a hybrid loss function combining 3D disentanglement loss and 3D pose loss is constructed to realize dual supervision at the bone and joint levels. Extensive experiments on Human3.6M and MPI-INF-3DHP datasets show that HADD achieves consistent performance improvement over state-of-the-art methods. On Human3.6M, it reduces average MPJPE by up to 12.9% compared with traditional disentangled-based methods, and outperforms mainstream diffusion-based probabilistic methods under lightweight inference configurations. On MPI-INF-3DHP, HADD achieves an MPJPE of 29.2 mm, a PCK of 98.5%, and an AUC of 78.1 under single-hypothesis inference, reaching new state-of-the-art performance on this dataset under the single-hypothesis inference setting with only ground-truth 2D joint sequences as input. This work provides an effective solution to the core limitations of existing 3D HPE methods, and also offers reliable technical support for fine-grained motion analysis and quantitative assessment in sports training and rehabilitation scenarios.
Related
- Diagnostic accuracy, time window estimation, and prognostic value of diffusion-weighted MRI in acute ischemic stroke: a systematic review and meta-analysis
- A Time-Dependent Diffusion MRI Framework for Clinical Characterisation of Human Brain Cellular Architecture
- Weighted estimation of multiple regions in CT-derived images of the pubic symphysis for age estimation using a deep learning framework with R-Mixup augmentation
- Glucose transport as a design constraint in white-zone meniscus tissue engineering: a reaction-diffusion framework
- Forecasting risk-factor distributions by sex and age, 1990-2050: a parametric estimation framework using GBD summary exposure values, demonstrated for Australia.
BibTeX
@article{jiang2026hadd,
title = {{HADD: A hierarchy-aware disentangled diffusion framework with spatio-temporal denoising for 3D human pose estimation}},
author = {Jinze Jiang and Zhongheng Jian and Hualing Zheng and Minsi Liang and Yuxing Shi and Liming Zhang},
journal = {PLOS One},
year = {2026},
month = sep,
doi = {10.1371/journal.pone.0359321},
url = {https://doi.org/10.1371/journal.pone.0359321},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - HADD: A hierarchy-aware disentangled diffusion framework with spatio-temporal denoising for 3D human pose estimation
AU - Jinze Jiang
AU - Zhongheng Jian
AU - Hualing Zheng
AU - Minsi Liang
AU - Yuxing Shi
AU - Liming Zhang
PY - 2026
DA - 2026/09/28
JO - PLOS One
DO - 10.1371/journal.pone.0359321
UR - https://doi.org/10.1371/journal.pone.0359321
AB - Monocular 3D Human Pose Estimation (3D HPE) is a core task in computer vision, which provides critical technical support for applications including VR/AR, immersive human-computer interaction, digital human animation, as well as sports motion capture, athletic technique diagnosis, physical fitness assessment and sports rehabilitation training. However, existing methods still face three key limitations: diffusion-based methods lack explicit human anatomical prior constraints in the diffusion process; disentangled-based methods suffer from severe hierarchical error accumulation along the skeletal kinematic tree; and most mainstream methods fail to fully mine the fine-grained hierarchical spatio-temporal correlations of the human skeleton, leading to increased estimation error for high-hierarchy joints. To address these challenges, we propose a novel Hierarchy-Aware Disentangled Diffusion framework with Spatio-Temporal Denoising (HADD) for monocular 3D HPE, which deeply integrates disentanglement strategy with the diffusion model and embeds skeletal hierarchical information into the full diffusion pipeline. In the forward diffusion process, HADD disentangles 3D pose into bone length and bone direction features, and injects Gaussian noise into the two features separately. In the reverse denoising process, we design a Hierarchical Spatial and Temporal Denoising (HSTD) module to accurately model the hierarchical spatio-temporal relationships of the human skeleton and alleviate error amplification of high-hierarchy joints. Meanwhile, a hybrid loss function combining 3D disentanglement loss and 3D pose loss is constructed to realize dual supervision at the bone and joint levels. Extensive experiments on Human3.6M and MPI-INF-3DHP datasets show that HADD achieves consistent performance improvement over state-of-the-art methods. On Human3.6M, it reduces average MPJPE by up to 12.9% compared with traditional disentangled-based methods, and outperforms mainstream diffusion-based probabilistic methods under lightweight inference configurations. On MPI-INF-3DHP, HADD achieves an MPJPE of 29.2 mm, a PCK of 98.5%, and an AUC of 78.1 under single-hypothesis inference, reaching new state-of-the-art performance on this dataset under the single-hypothesis inference setting with only ground-truth 2D joint sequences as input. This work provides an effective solution to the core limitations of existing 3D HPE methods, and also offers reliable technical support for fine-grained motion analysis and quantitative assessment in sports training and rehabilitation scenarios.
SN - 1932-6203
ER -