Research map / 3D vision / Human pose estimation
PSVT: End-to-End Multi-Person 3D Pose and Shape Estimation With Progressive Video Transformers
Zhongwei Qiu, Qiansheng Yang, Jian Wang, Haocheng Feng, Junyu Han, Errui Ding, Chang Xu, Dongmei Fu, Jingdong Wang
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 progressive decoding with PGA rather than Split-PoSh (ablation) | PSVT · 3DPW | PA-MPJPE: 46.9 -> 43.5 | Improved |
| use progressive decoding with PGA rather than Split-PoSh (ablation) | PSVT · 3DPW | MPJPE: 78.3 -> 73.1 | Improved |
| use progressive decoding with PGA rather than Split-PoSh (ablation) | PSVT · 3DPW | MPVE: 87.5 -> 84.0 | Improved |
| add PGA instead of a convolutional decoder (ablation) | image-encoder PSVT · 3DPW | PA-MPJPE: 49.8 -> 45.7 | Improved |
| add PGA instead of a convolutional decoder (ablation) | image-encoder PSVT · 3DPW | MPJPE: 86.2 -> 75.5 | Improved |
| add PGA instead of a convolutional decoder (ablation) | image-encoder PSVT · 3DPW | MPVE: 96.9 -> 84.9 | Improved |
| add a spatio-temporal encoder rather than a spatial encoder (ablation) | PSVT with PGA · 3DPW | PA-MPJPE: 45.7 -> 44.1 | Improved |
| add a spatio-temporal encoder rather than a spatial encoder (ablation) | PSVT with PGA · 3DPW | MPJPE: 75.5 -> 74.2 | Improved |
| add a spatio-temporal encoder rather than a spatial encoder (ablation) | PSVT with PGA · 3DPW | MPVE: 84.9 -> 84.5 | Improved |
| add Token Aligning to STSD (ablation) | PSVT · 3DPW | MPJPE: 76.6 -> 75.5 | Improved |
| add Window Pose Self-Attention to STSD (ablation) | PSVT · 3DPW | MPJPE: 77.1 -> 75.5 | Improved |
| add Window Shape Self-Attention to STSD (ablation) | PSVT · 3DPW | MPJPE: 76.1 -> 75.5 | Improved |
Later papers that compared with this method
2 later accepted papers report PSVT as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- SAT-HMR: Real-Time Multi-Person 3D Mesh Estimation via Scale-Adaptive Tokens (CVPR 2025 · better on 1 of 1 shared result)
- Multi-HMR: Multi-Person Whole-Body Human Mesh Recovery in a Single Shot (ECCV 2024 · better on 3 of 3 shared results)
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
- End-to-End Multi-Person Pose Estimation with Pose-Aware Video Transformer (AAAI 2026 · similarity 0.90)
- SVTformer: Spatial-View-Temporal Transformer for Multi-View 3D Human Pose Estimation (AAAI 2025 · similarity 0.89)
- Global-to-Local Modeling for Video-Based 3D Human Pose and Shape Estimation (CVPR 2023 · similarity 0.88)
- Direct Multi-view Multi-person 3D Pose Estimation (NeurIPS 2021 · similarity 0.87)
- ExtPose: Robust and Coherent Pose Estimation by Extending ViTs (ICML 2025 · similarity 0.86)
- 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.
