Research map / 3D vision / Human pose estimation
MANIKIN: Biomechanically Accurate Neural Inverse Kinematics for Human Motion Estimation
Jiaxi Jiang, Paul Streli, Xuejing Luo, Christoph Gebhardt, Christian Holz
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add analytic upper-body IK (ablation) | MANIKIN · AMASS MPI | MPJPE (cm): 7.20 -> 6.41 | Improved |
| add analytic lower-body IK (ablation) | MANIKIN · AMASS MPI | MPJPE (cm): 7.06 -> 6.41 | Improved |
| use analytic IK instead of optimization-based IK (ablation) | MANIKIN · AMASS MPI | MPJPE (cm): 7.65 -> 6.41 | Improved |
| use analytic IK instead of learning-based IK (ablation) | MANIKIN · AMASS MPI | MPJPE (cm): 7.96 -> 6.41 | Improved |
| add mid-joint loss (ablation) | MANIKIN · AMASS MPI | MPJPE (cm): 6.59 -> 6.41 | Improved |
| add swivel angle loss (ablation) | MANIKIN · AMASS MPI | MPJPE (cm): 6.62 -> 6.41 | Improved |
| predict swivel angle as continuous [cos, sin] instead of a single value (ablation) | MANIKIN · AMASS MPI | MPJPE (cm): 7.14 -> 6.41 | Improved |
| predict swivel angle with network instead of traditional search (ablation) | MANIKIN · AMASS MPI | MPJPE (cm): 8.82 -> 6.41 | Improved |
| use biomechanical limb model with swivel IK instead of SMPL for 5-point input | Transformer / RNN / MLP backbones · 5-point input MPI | MPJPE (cm): 8.68 -> 5.37 (MLP) | Improved |
| MANIKIN-L vs AvatarJLM on speed | MANIKIN-L · AMASS | FPS: 31.1 -> 30.5 | No clear effect |
| MANIKIN-S vs AvatarJLM in speed | MANIKIN-S · AMASS | FPS: 31.1 -> 580 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- AvatarPoser* — AvatarPoser: Articulated Full-Body Pose Tracking from Sparse Motion Sensing (ECCV 2022) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- HybrIK: A Hybrid Analytical-Neural Inverse Kinematics Solution for 3D Human Pose and Shape Estimation (CVPR 2021 · similarity 0.88)
- Reconstructing Humans with a Biomechanically Accurate Skeleton (CVPR 2025 · similarity 0.88)
- MS-MANO: Enabling Hand Pose Tracking with Biomechanical Constraints (CVPR 2024 · similarity 0.86)
- Optimal-state Dynamics Estimation for Physics-based Human Motion Capture from Videos (NeurIPS 2024 · similarity 0.86)
- CoDA: Coordinated Diffusion Noise Optimization for Whole-Body Manipulation of Articulated Objects (NeurIPS 2025 · similarity 0.86)
- Forecasting of 3D Whole-body Human Poses with Grasping Objects (CVPR 2024 · similarity 0.85)
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.
