Research map / 3D vision / Human pose estimation
KineST: A Kinematics-guided Spatiotemporal State Space Model for Human Motion Tracking from Sparse Signals
Shuting Zhao, Zeyu Xiao, Xinrong Chen
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 |
|---|---|---|---|
| use unified kinematic scan (UKS) instead of SMPL index-order scan (ablation) | KineST · AMASS Protocol 1 | MPJPE (cm): 3.11 -> 2.86 | Improved |
| use unified kinematic scan (UKS) instead of five-branch kinematic scan (FKS) (ablation) | KineST · AMASS Protocol 1 | Jitter: 7.01 -> 5.97 | Improved |
| add kinematic scanning (FKS) over index-order scan (ablation) | KineST · AMASS Protocol 1 | MPJVE (cm/s): 17.81 -> 16.25 | Improved |
| use spatiotemporal mixing mechanism (STMM) instead of pure temporal modeling in SKFM (ablation) | KineST · AMASS Protocol 1 | Jitter: 7.83 -> 5.97 | Improved |
| use STMM instead of holistic pure spatial modeling in SKFM (ablation) | KineST · AMASS Protocol 1 | MPJVE (cm/s): 16.77 -> 15.26 | Improved |
| use STMM instead of token-wise spatial modeling in SKFM (ablation) | KineST · AMASS Protocol 1 | MPJRE / Jitter: MPJRE 2.23 -> 2.25; Jitter 9.31 -> 5.97 | Mixed |
| add geometric angular velocity loss (ablation) | KineST · AMASS Protocol 1 | MPJVE (cm/s): 16.10 -> 15.26 | Improved |
| add finite-difference angular velocity loss (ablation) | KineST · AMASS Protocol 1 | MPJPE (cm): 2.87 -> 3.03 | Hurt |
| use geometric instead of finite-difference angular velocity loss (ablation) | KineST · AMASS Protocol 1 | MPJRE (deg): 2.29 -> 2.25 | Improved |
| KineST versus MMD, previous best SSM method | KineST · AMASS Protocol 1 | MPJPE (cm): 3.22 -> 2.86 | Improved |
| KineST compared with larger models AvatarJLM and SAGE on position error | KineST · AMASS Protocol 2 | MPJPE (cm): AvatarJLM 4.93 vs KineST 5.17 | No clear effect |
Closest papers by meaning
- AvatarPoser: Articulated Full-Body Pose Tracking from Sparse Motion Sensing (ECCV 2022 · similarity 0.88)
- HumMUSS: Human Motion Understanding using State Space Models (CVPR 2024 · similarity 0.86)
- Optimal-state Dynamics Estimation for Physics-based Human Motion Capture from Videos (NeurIPS 2024 · similarity 0.86)
- HMD-Poser: On-Device Real-time Human Motion Tracking from Scalable Sparse Observations (CVPR 2024 · similarity 0.85)
- A Unified Diffusion Framework for Scene-aware Human Motion Estimation from Sparse Signals (CVPR 2024 · similarity 0.85)
- MV-SSM: Multi-View State Space Modeling for 3D Human Pose Estimation (CVPR 2025 · 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.
