Research map / Point clouds / Point cloud classification
Dynamic Mesh Recovery from Partial Point Cloud Sequence
Hojun Jang, Minkwan Kim, Jinseok Bae, Young Min Kim
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 auxiliary pose loss to kinematics learner (ablation) | Kinematics learner · AMASS subset | vertex error: 0.0309 -> 0.0215 | Improved |
| add volume fitting loss to kinematics learner (ablation) | Kinematics learner · AMASS subset | vertex error: 0.1217 -> 0.0215 | Improved |
| pretrain kinematics prior then use feature follower instead of training directly on partial input (ablation) | Feature follower pipeline · Synthetic human motion | vertex error: 0.1018 -> 0.0132 | Improved |
| add feature following loss L_F (ablation) | Feature follower · Synthetic human motion | vertex error: 0.0166 -> 0.0132 | Improved |
| use full-capacity transformer in feature follower instead of quartered (ablation) | Feature follower · Synthetic human motion | vertex error: 0.0190 -> 0.0132 | Improved |
| sequence-level temporal modeling versus per-frame baselines | Ours vs VoteHMR · Synthetic human motion | vertex error: 0.0258 -> 0.0132 | Improved |
| Ours vs optimization-based Zuo et al. on real data | Ours · Berkeley MHAD | Chamfer distance: 0.0014 (Zuo) vs 0.0027 (Ours) | Hurt |
Closest papers by meaning
- Uncertainty-Aware Human Mesh Recovery From Video by Learning Part-Based 3D Dynamics (ICCV 2021 · similarity 0.90)
- HuMoR: 3D Human Motion Model for Robust Pose Estimation (ICCV 2021 · similarity 0.89)
- Contextually Plausible and Diverse 3D Human Motion Prediction (ICCV 2021 · similarity 0.88)
- Learning 3D Human Dynamics From Video (CVPR 2019 · similarity 0.88)
- DPMesh: Exploiting Diffusion Prior for Occluded Human Mesh Recovery (CVPR 2024 · similarity 0.87)
- VideoMDM: Towards 3D Human Motion Generation From 2D Supervision (NeurIPS 2026 · similarity 0.86)
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.
