mlsanityData to 14 Oct 2026

Efficient 2D to Full 3D Human Pose Uplifting including Joint Rotations

arXiv 2504.09953 ↗First posted 14 Apr 2025cs.CVcited by 1

Not on the map. Why?

Abstract (arXiv)

In sports analytics, accurately capturing both the 3D locations and rotations of body joints is essential for understanding an athlete's biomechanics. While Human Mesh Recovery (HMR) models can estimate joint rotations, they often exhibit lower accuracy in joint localization compared to 3D Human Pose Estimation (HPE) models. Recent work addressed this limitation by combining a 3D HPE model with inverse kinematics (IK) to estimate both joint locations and rotations. However, IK is computationally expensive. To overcome this, we propose a novel 2D-to-3D uplifting model that directly estimates 3D human poses, including joint rotations, in a single forward pass. We investigate multiple rotation representations, loss functions, and training strategies - both with and without access to ground truth rotations. Our models achieve state-of-the-art accuracy in rotation estimation, are 150 times faster than the IK-based approach, and surpass HMR models in joint localization precision.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 24 of its 29 references that are in the map. First posted on arXiv 14 Apr 2025.

Builds on

24 of 29 references.

References · 24

  1. 2019Expressive Body Capture: 3D Hands, Face, and Body from a Single Imagecited by 1,612
  2. 2019AMASS: Archive of Motion Capture as Surface Shapescited by 1,236
  3. 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,214
  4. 2017PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenescited by 775
  5. 20183D human pose estimation in video with temporal convolutions and semi-supervised trainingcited by 403
  6. 2023Humans in 4D: Reconstructing and Tracking Humans with Transformerscited by 350
  7. 20213D Human Pose Estimation with Spatial and Temporal Transformerscited by 272
  8. 2016DeMoN: Depth and Motion Network for Learning Monocular Stereocited by 201
  9. 2020Monocular Expressive Body Regression through Body-Driven Attentioncited by 155
  10. 2023SMPLer-X: Scaling Up Expressive Human Pose and Shape Estimationcited by 155
  11. 2021Putting People in their Place: Monocular Regression of 3D People in Depthcited by 142
  12. 2021Collaborative Regression of Expressive Bodies using Moderationcited by 137
12 more
  1. 2024Multi-HMR: Multi-Person Whole-Body Human Mesh Recovery in a Single Shotcited by 101
  2. 2021Exploiting Temporal Contexts with Strided Transformer for 3D Human Pose Estimationcited by 94
  3. 2020An Analysis of SVD for Deep Rotation Estimationcited by 86
  4. 2020Accurate 3D Hand Pose Estimation for Whole-Body 3D Human Mesh Estimationcited by 81
  5. 2021Learning to Estimate Robust 3D Human Mesh from In-the-Wild Crowded Scenescited by 65
  6. 2021Conditional Directed Graph Convolution for 3D Human Pose Estimationcited by 52
  7. 2022Uplift and Upsample: Efficient 3D Human Pose Estimation with Uplifting Transformerscited by 33
  8. 2023PSVT: End-to-End Multi-person 3D Pose and Shape Estimation with Progressive Video Transformerscited by 32
  9. 2018Deep-6DPose: Recovering 6D Object Pose from a Single RGB Imagecited by 26
  10. 2018Occlusion Resistant Object Rotation Regression from Point Cloud Segmentscited by 9
  11. 2023Loss it right: Euclidean and Riemannian Metrics in Learning-based Visual Odometrycited by 6
  12. 2024Leveraging Anthropometric Measurements to Improve Human Mesh Estimation and Ensure Consistent Body Shapescited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026DuetHOI: Language-Guided Bimanual Hand--Object Motion Generation with Articulation Planning and Contact Refinementcited by 0