Research map / 3D vision / Human pose estimation
PARE: Part Attention Regressor for 3D Human Body Estimation
Muhammed Kocabas, Chun-Hao P. Huang, Otmar Hilliges, Michael J. Black
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 |
|---|---|---|---|
| supervise attention branch with part segmentation instead of none (ablation) | PARE · 3DPW-OCC | PA-MPJPE: 62.7 -> 61.2 | Improved |
| use part supervision then drop it (mixed) instead of part supervision throughout (ablation) | PARE · 3DPW | MPJPE: 94.5 -> 93.4 | Improved |
| use mixed supervision instead of part supervision throughout (ablation) | PARE · 3DPW-OCC | PA-MPJPE: 61.2 -> 61.6 | No clear effect |
| soft attention instead of pooling with part supervision (ablation) | PARE · 3DPW-OCC | PA-MPJPE: 64.8 -> 61.2 | Improved |
| soft attention instead of pooling with joint supervision (ablation) | PARE · 3DPW-OCC | MPJPE: 95.4 -> 98.9 | Hurt |
| part segmentation instead of joint supervision with soft attention (ablation) | PARE · 3DPW-OCC | PA-MPJPE: 63.9 -> 61.2 | Improved |
| add SynthOcc augmentation (ablation) | PARE · 3DPW | MPJPE: 95.0 -> 94.5 | Improved |
| add random crop on top of SynthOcc (ablation) | PARE · 3DPW-OCC | PA-MPJPE: 61.2 -> 62.6 | Hurt |
| use HRNet-W32 instead of ResNet-50 (ablation) | PARE · 3DPW | PA-MPJPE: 57.1 -> 54.3 | Improved |
| replace HMR-EFT with PARE (ResNet-50) | HMR-EFT · 3DPW | PA-MPJPE: 59.9 -> 57.1 | Improved |
| replace HMR-EFT with PARE (HRNet-W32) | HMR-EFT · 3DPW | PA-MPJPE: 55.9 -> 54.3 | Improved |
| replace SPIN with PARE | SPIN · 3DOH | PA-MPJPE: 68.3 -> 44.3 | 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.
- SPIN — SPIN: Structure-Preserving Inner Offset Network for Scene Text Recognition (AAAI 2021) · better on 6 of 6 shared results
- VIBE — VIBE: Video Inference for Human Body Pose and Shape Estimation (CVPR 2020) · 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
- Learning To Regress Bodies From Images Using Differentiable Semantic Rendering (ICCV 2021 · similarity 0.87)
- Learning Visibility for Robust Dense Human Body Estimation (ECCV 2022 · similarity 0.87)
- Attention Mechanism Exploits Temporal Contexts: Real-Time 3D Human Pose Reconstruction (CVPR 2020 · similarity 0.87)
- Towards Part-aware Monocular 3D Human Pose Estimation: An Architecture Search Approach (ECCV 2020 · similarity 0.86)
- FinePOSE: Fine-Grained Prompt-Driven 3D Human Pose Estimation via Diffusion Models (CVPR 2024 · similarity 0.86)
- HiPoser: 3D Human Pose Estimation with Hierarchical Shared Learning at Parts-Level Using Inertial Measurement Units (AAAI 2025 · 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.
