mlsanityData to 14 Oct 2026

Normalizing Flows on the Product Space of SO(3) Manifolds for Probabilistic Human Pose Modeling

arXiv 2404.05675 ↗First posted 8 Apr 2024cs.CVcited by 7

Not on the map. Why?

Abstract (arXiv)

Normalizing flows have proven their efficacy for density estimation in Euclidean space, but their application to rotational representations, crucial in various domains such as robotics or human pose modeling, remains underexplored. Probabilistic models of the human pose can benefit from approaches that rigorously consider the rotational nature of human joints. For this purpose, we introduce HuProSO3, a normalizing flow model that operates on a high-dimensional product space of SO(3) manifolds, modeling the joint distribution for human joints with three degrees of freedom. HuProSO3's advantage over state-of-the-art approaches is demonstrated through its superior modeling accuracy in three different applications and its capability to evaluate the exact likelihood. This work not only addresses the technical challenge of learning densities on SO(3) manifolds, but it also has broader implications for domains where the probabilistic regression of correlated 3D rotations is of importance.

Key facts. Cited by 7 papers in the map (data through 14 Oct 2026). Builds on 28 of its 41 references that are in the map. First posted on arXiv 8 Apr 2024.

Builds on

28 of 41 references.

References · 28

  1. 2016Density estimation using Real NVPcited by 2,042
  2. 2019Expressive Body Capture: 3D Hands, Face, and Body from a Single Imagecited by 1,612
  3. 2018Glow: Generative Flow with Invertible 1x1 Convolutionscited by 1,545
  4. 2019Normalizing Flows for Probabilistic Modeling and Inferencecited by 1,323
  5. 2019AMASS: Archive of Motion Capture as Surface Shapescited by 1,236
  6. 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,214
  7. 2014NICE: Non-linear Independent Components Estimationcited by 1,058
  8. 2016Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Imagecited by 695
  9. 2017Masked Autoregressive Flow for Density Estimationcited by 684
  10. 2019Neural Spline Flowscited by 596
  11. 2021HuMoR: 3D Human Motion Model for Robust Pose Estimationcited by 324
  12. 2015MADE: Masked Autoencoder for Distribution Estimationcited by 307
16 more
  1. 2021Probabilistic Modeling for Human Mesh Recoverycited by 153
  2. 2020Weakly Supervised 3D Human Pose and Shape Reconstruction with Normalizing Flowscited by 94
  3. 2022Pose-NDF: Modeling Human Pose Manifolds with Neural Distance Fieldscited by 85
  4. 2021Implicit-PDF: Non-Parametric Representation of Probability Distributions on the Rotation Manifoldcited by 65
  5. 2022GFPose: Learning 3D Human Pose Prior with Gradient Fieldscited by 57
  6. 2023normflows: A PyTorch Package for Normalizing Flowscited by 56
  7. 2021Hierarchical Kinematic Probability Distributions for 3D Human Shape and Pose Estimation from Images in the Wildcited by 48
  8. 2021Probabilistic 3D Human Shape and Pose Estimation from Multiple Unconstrained Images in the Wildcited by 40
  9. 2021Adversarial Parametric Pose Priorcited by 38
  10. 2022Motron: Multimodal Probabilistic Human Motion Forecastingcited by 30
  11. 2023HuManiFlow: Ancestor-Conditioned Normalising Flows on SO(3) Manifolds for Human Pose and Shape Distribution Estimationcited by 24
  12. 2020Risk-Sensitive Sequential Action Control with Multi-Modal Human Trajectory Forecasting for Safe Crowd-Robot Interactioncited by 21
  13. 2021ProtoRes: Proto-Residual Network for Pose Authoring via Learned Inverse Kinematicscited by 17
  14. 2023Delving into Discrete Normalizing Flows on SO(3) Manifold for Probabilistic Rotation Modelingcited by 14
  15. 2022SMPL-IK: Learned Morphology-Aware Inverse Kinematics for AI Driven Artistic Workflowscited by 6
  16. 2020Unifying supervised learning and VAEs -- coverage, systematics and goodness-of-fit in normalizing-flow based neural network models for astro-particle reconstructionscited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 1; 2025: 4; 2026: 2.

7 papers.

  1. 2025DPoser-X: Diffusion Model as Robust 3D Whole-body Human Pose Priorcited by 13
  2. 2024Utilizing Uncertainty in 2D Pose Detectors for Probabilistic 3D Human Mesh Recoverycited by 4
  3. 2025Neural Human Pose Priorcited by 0
  4. 2025LieHMR: Autoregressive Human Mesh Recovery with SO(3) Diffusioncited by 0
  5. 2025BabyFlow: 3D modeling of realistic and expressive infant facescited by 0
  6. 2026Learning Manifold and Ito Dynamics with Branched Neural Rough Differential Equationscited by 0
  7. 2026Naturalness-guided Manifold Flow Matching for Sign Language Productioncited by 0