Research map / 3D vision / Human pose estimation
Monocular, One-Stage, Regression of Multiple 3D People
Yu Sun, Qian Bao, Wu Liu, Yili Fu, Michael J. Black, Tao Mei
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 collision-aware representation (CAR, gamma=0.2) (ablation) | ROMP · 3DPW-PC | PMPJPE (mm): 79.7 -> 75.8 | Improved |
| add CAR (ablation) | ROMP · Crowdpose test | AP0.5: 54.1 -> 59.7 | Improved |
| add CAR (ablation) | ROMP · Crowdpose validation | AP0.5: 55.6 -> 58.6 | Improved |
| add CAR (gamma=0.2) (ablation) | ROMP · 3DPW-NC | PMPJPE (mm): 56.7 -> 57.1 | Hurt |
| add CAR (gamma=0.2) (ablation) | ROMP · 3DPW-OC | PMPJPE (mm): 67.0 -> 67.1 | No clear effect |
| use repulsion coefficient gamma=0.1 instead of 0.2 (ablation) | ROMP+CAR · 3DPW-PC | PMPJPE (mm): 75.8 -> 77.6 | Hurt |
| use repulsion coefficient gamma=0.3 instead of 0.2 (ablation) | ROMP+CAR · 3DPW-PC | PMPJPE (mm): 75.8 -> 77.0 | Hurt |
| use one-stage ROMP instead of multi-stage YOLO+VIBE on same backbone | ROMP · 3DPW (Protocol 1) | MPJPE (mm): 94.7 -> 87.0 | Improved |
| train with extra datasets | ROMP (ResNet-50) · 3DPW (Protocol 1) | MPJPE (mm): 87.0 -> 80.1 | Improved |
| use HRNet-32 instead of ResNet-50 backbone | ROMP (extra data) · 3DPW (Protocol 3) | PMPJPE (mm): 49.7 -> 47.3 | Improved |
| use ROMP instead of CRMH | ROMP · CMU Panoptic | mean MPJPE (mm): 143.2 -> 134.6 | Improved |
| use one-stage ROMP instead of VIBE for speed | ROMP (ResNet-50) · Runtime | FPS: 10.9 -> 30.9 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- VIBE — VIBE: Video Inference for Human Body Pose and Shape Estimation (CVPR 2020) · better on 2 of 2 shared results
- SPIN — SPIN: Structure-Preserving Inner Offset Network for Scene Text Recognition (AAAI 2021) · 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
- Multi-HMR: Multi-Person Whole-Body Human Mesh Recovery in a Single Shot (ECCV 2024 · similarity 0.89)
- Recovering 3D Human Mesh From Monocular Images: A Survey (TPAMI 2023 · similarity 0.89)
- Body Meshes as Points (CVPR 2021 · similarity 0.88)
- Multi-Person 3D Pose and Shape Estimation via Inverse Kinematics and Refinement (ECCV 2022 · similarity 0.88)
- SAT-HMR: Real-Time Multi-Person 3D Mesh Estimation via Scale-Adaptive Tokens (CVPR 2025 · similarity 0.88)
- SAM 3D Body: Robust Full-Body Human Mesh Recovery (CVPR 2026 · 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.
