Research map / 3D vision / Human pose estimation
Learning To Estimate Robust 3D Human Mesh From In-the-Wild Crowded Scenes
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 2D pose guidance to the image feature (ablation) | 3DCrowdNet · 3DPW-Crowd | MPJPE: 109.6 -> 85.8 | Helped |
| Add 2D pose guidance to the image feature (ablation) | 3DCrowdNet · 3DPW-Crowd | PA-MPJPE: 63.3 -> 55.8 | Helped |
| Replace the SPIN-style global-pooling regressor with the joint-based regressor (ablation) | 3DCrowdNet · 3DPW-Crowd | MPJPE: 89.0 -> 85.8 | Helped |
| Replace the SPIN-style global-pooling regressor with the joint-based regressor (ablation) | 3DCrowdNet · 3DPW-Crowd | PA-MPJPE: 59.5 -> 55.8 | Helped |
| Keep feature sampling at joint points instead of expanding it to the whole feature map (ablation) | 3DCrowdNet joint-based regressor · 3DPW-Crowd | MPJPE: 89.1 -> 85.8 | Helped |
| Keep feature sampling at joint points instead of expanding it to a 5-by-5 grid (ablation) | 3DCrowdNet joint-based regressor · 3DPW-Crowd | MPJPE: 88.2 -> 85.8 | Helped |
| Estimate a 3D pose rather than a 2D pose as the intermediate target (ablation) | 3DCrowdNet joint-based regressor · 3DPW-Crowd | MPJPE: 88.3 -> 85.8 | Helped |
| Estimate a 3D pose rather than a 2D pose as the intermediate target (ablation) | 3DCrowdNet joint-based regressor · 3DPW-Crowd | PA-MPJPE: 56.4 -> 55.8 | Helped |
| Use 3DCrowdNet* instead of ROMP* | 3DPW-Crowd · 3DPW-Crowd | MPJPE: 104.8 -> 85.8 | Helped |
| Use 3DCrowdNet instead of ROMP | CMU-Panoptic · CMU-Panoptic | MPJPE: 134.6 -> 127.6 | Helped |
| Use 3DCrowdNet instead of Fang et al. | 3DPW · 3DPW | MPJPE: 85.1 -> 81.7 | Helped |
Closest papers by meaning
- Benchmarking and Analyzing 3D Human Pose and Shape Estimation Beyond Algorithms (NeurIPS 2022 · similarity 0.89)
- TokenHMR: Advancing Human Mesh Recovery with a Tokenized Pose Representation (CVPR 2024 · similarity 0.88)
- Sim2real transfer learning for 3D human pose estimation: motion to the rescue (NeurIPS 2019 · similarity 0.88)
- FreeMan: Towards Benchmarking 3D Human Pose Estimation under Real-World Conditions (CVPR 2024 · similarity 0.88)
- Reconstructing People, Places, and Cameras (CVPR 2025 · similarity 0.88)
- SAT-HMR: Real-Time Multi-Person 3D Mesh Estimation via Scale-Adaptive Tokens (CVPR 2025 · 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.