Research map / 3D vision / Human pose estimation
PS-Mamba: Spatial-Temporal Graph Mamba for Pose Sequence Refinement
Haoye Dong, Gim Hee Lee
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 ST-GSS block (ablation) | PS-Mamba · Human3.6M 3D | MPJPE (mm): 370.23 -> 49.60 | Improved |
| add ST-SSM (ablation) | PS-Mamba · Human3.6M 3D | MPJPE (mm): 51.97 -> 49.60 | Improved |
| add ST-SSM (ablation) | PS-Mamba · Human3.6M 3D | Accel (mm/s2): 1.25 -> 0.92 | Improved |
| add ST-Graph (ablation) | PS-Mamba · Human3.6M 3D | MPJPE (mm): 50.75 -> 49.60 | Improved |
| add dynamic graph weight matrix W (ablation) | PS-Mamba · Human3.6M 3D | MPJPE (mm): 50.74 -> 49.60 | Improved |
| add temporal strategy (ablation) | PS-Mamba · Human3.6M 3D | MPJPE (mm): 50.27 -> 49.60 | Improved |
| add residual connections (ablation) | PS-Mamba · Human3.6M 3D | MPJPE (mm): 50.21 -> 49.60 | Improved |
| add MPJPE loss (ablation) | PS-Mamba · Human3.6M 3D | MPJPE (mm): 446.50 -> 49.60 | Improved |
| add PA-MPJPE loss (ablation) | PS-Mamba · Human3.6M 3D | MPJPE (mm): 51.02 -> 49.60 | Improved |
| add acceleration loss (ablation) | PS-Mamba · Human3.6M 3D | Accel (mm/s2): 0.95 -> 0.92 | Improved |
| increase window size from 4 to 40 (ablation) | PS-Mamba · Human3.6M FCN input | MPJPE (mm): 53.26 -> 50.13 | Improved |
| increase window size from 40 to 50 (ablation) | PS-Mamba · Human3.6M FCN input | MPJPE (mm): 50.13 -> 51.91 | Hurt |
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.
- SynSP — SynSP: Synergy of Smoothness and Precision in Pose Sequences Refinement (CVPR 2024) · better on 2 of 2 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
- PoseMamba: Monocular 3D Human Pose Estimation with Bidirectional Global-Local Spatio-Temporal State Space Model (AAAI 2025 · similarity 0.88)
- Pose Magic: Efficient and Temporally Consistent Human Pose Estimation with a Hybrid Mamba-GCN Network (AAAI 2025 · similarity 0.88)
- GRPose: Learning Graph Relations for Human Image Generation with Pose Priors (AAAI 2025 · similarity 0.87)
- Spatial-Mamba: Effective Visual State Space Models via Structure-Aware State Fusion (ICLR 2025 · similarity 0.87)
- Video-Based Human Pose Regression via Decoupled Space-Time Aggregation (CVPR 2024 · similarity 0.86)
- Vision Mamba Mender (NeurIPS 2024 · 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.
