Research map / 3D vision / 3D generation
MotionMAR: Multi-scale Auto-Regressive Human Motion Reconstruction from Sparse Observations
Yuhua Luo, Junsheng Zhang, Mengyin Liu, Xincheng Lin, Ming Yan, Zhudi Chen, Chenglu Wen, Lan Xu, Siqi Shen, Cheng Wang
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 Temporal Multi-scale Tokenization (ablation) | MotionMAR · AMASS S1 | MPJPE: 3.41 -> 2.82 | Improved |
| add Motion Refinement Network (Bi-GRU) (ablation) | MotionMAR · AMASS S1 | Jitter: 42.85 -> 5.17 | Improved |
| add Motion Refinement Network (Bi-GRU) (ablation) | MotionMAR · AMASS S1 | MPJVE: 43.9 -> 16.23 | Improved |
| add Scale-aware Control module (ablation) | MotionMAR · AMASS S1 | Root PE: 3.22 -> 2.58 | Improved |
| add Scale-aware Control module (ablation) | MotionMAR · AMASS S1 | Hand PE: 1.21 -> 0.83 | Improved |
| add MRN (ablation) | MotionMAR · AMASS S1 | Hand PE: 0.84 -> 0.83 | No clear effect |
| replace SAGE single-scale VQ with TMT (ablation) | SAGE · AMASS S1 | reconstruction error | Improved |
| MotionMAR versus RPM on jitter in large-scale setting | MotionMAR · AMASS S3 | Jitter: 3.67 (RPM) vs 4.14 | Hurt |
Closest papers by meaning
- Natural Human Motion Recovery by Aligning High-Order Temporal Dynamics from Monocular Videos (CVPR 2026 · similarity 0.90)
- WaveAR: Wavelet-Aware Continuous Autoregressive Diffusion for Accurate Human Motion Prediction (NeurIPS 2025 · similarity 0.89)
- A Unified Diffusion Framework for Scene-aware Human Motion Estimation from Sparse Signals (CVPR 2024 · similarity 0.88)
- Reconstructing People, Places, and Cameras (CVPR 2025 · similarity 0.88)
- HumanMM: Global Human Motion Recovery from Multi-shot Videos (CVPR 2025 · similarity 0.87)
- RoHM: Robust Human Motion Reconstruction via Diffusion (CVPR 2024 · similarity 0.87)
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.
