Research map / 3D vision / Human pose estimation
HuManiFlow: Ancestor-Conditioned Normalising Flows on SO(3) Manifolds for Human Pose and Shape Distribution Estimation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Factorize pose into ancestor-conditioned body-part rotations and model the rotations on SO(3) with LRS flows rather than a concatenated full-body axis-angle flow (ablation) | HuManiFlow pose distribution · 3DPW | MPJPE-PA (sample min.): 46.1 -> 39.9 | Helped |
| Push ancestor-conditioned axis-angle distributions onto SO(3) (ablation) | Ancestor-conditioned LRS-NF · 3DPW | MPJPE-PA (sample min.): 41.1 -> 39.9 | Helped |
| Use ancestor-conditioned SO(3) LRS-NFs instead of the matrix-Fisher distribution (ablation) | Ancestor-conditioned SO(3) pose distributions · 3DPW | MPJPE-PA (sample min.): 43.4 -> 39.9 | Helped |
| Add the visibility-masked 2DKP samples loss (ablation) | HuManiFlow trained with NLL · 3DPW | 2DKP Error (samples): 8.1 -> 6.2 | Helped |
| Add non-probabilistic 3D keypoint and vertex point-estimate losses (ablation) | HuManiFlow trained with NLL and 2DKP samples loss · 3DPW | 3DKP Spread (visible): 42.8 -> 39.3 | Hurt |
| Add non-probabilistic 3D keypoint and vertex point-estimate losses (ablation) | HuManiFlow trained with NLL and 2DKP samples loss · 3DPW Cropped | MPJPE-PA (point): 78.2 -> 83.4 | Hurt |
| Use HuManiFlow instead of HierProbHuman | Probabilistic pose estimation · 3DPW | MPJPE-PA (sample min.): 43.8 -> 39.9 | Helped |
| Use HuManiFlow instead of HierProbHuman | Probabilistic shape estimation · SSP-3D | PVE-T-SC (sample min.): 8.7 -> 8.3 | Helped |
| Use HuManiFlow with an image-conditioned prior for model fitting rather than ProHMR with its image-conditioned prior | SMPL model fitting · 3DPW | MPJPE-PA: 55.1 -> 51.2 | Helped |
Closest papers by meaning
- Probabilistic 3D Human Shape and Pose Estimation From Multiple Unconstrained Images in the Wild (CVPR 2021 · similarity 0.90)
- Normalizing Flows on the Product Space of SO(3) Manifolds for Probabilistic Human Pose Modeling (CVPR 2024 · similarity 0.90)
- Capturing the Motion of Every Joint: 3D Human Pose and Shape Estimation with Independent Tokens (ICLR 2023 · similarity 0.90)
- ManiPose: Manifold-Constrained Multi-Hypothesis 3D Human Pose Estimation (NeurIPS 2024 · similarity 0.90)
- Self-Supervised 3D Human Mesh Recovery from a Single Image with Uncertainty-Aware Learning (AAAI 2024 · similarity 0.88)
- ODHSR: Online Dense 3D Reconstruction of Humans and Scenes from Monocular Videos (CVPR 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.