Research map / 3D vision / Human pose estimation
UST-Hand: An Uncertainty-aware Spatiotemporal Point Cloud Interaction Network for 3D Self-supervised Hand Pose Estimation
Tianhao Han, Haoyang Zhang, Liang Xie, Haochen Chang, Kun Gao, Yuan Cheng, Pengfei Ren, Erwei Yin
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 heatmap-based confidence module (ablation) | UST-Hand · HanCo | MPVPE: 6.42 -> 5.82 | Improved |
| add heatmap-based confidence module (ablation) | UST-Hand · OakInk-MV | MPVPE: 11.75 -> 10.02 | Improved |
| add projection fusion module (ablation) | UST-Hand · DexYCB-MV | MPVPE: 8.70 -> 8.16 | Improved |
| add projection fusion module (ablation) | UST-Hand · OakInk-MV | MPVPE: 10.41 -> 10.02 | Improved |
| add Spatiotemporal Point Transformer (ablation) | UST-Hand · DexYCB-MV | MPVPE: 8.78 -> 8.16 | Improved |
| add Spatiotemporal Point Transformer (ablation) | UST-Hand · HanCo | MPVPE: 6.05 -> 5.82 | Improved |
| add confidence-aware self-attention (ablation) | UST-Hand · DexYCB-MV | MPVPE: 8.28 -> 8.16 | Improved |
| add adaptive GCN (ablation) | UST-Hand · DexYCB-MV | MPVPE: 8.22 -> 8.16 | Improved |
| add temporal attention in STPT (ablation) | UST-Hand · DexYCB-MV | MPVPE: 8.55 -> 8.16 | Improved |
| increase temporal window from 1 to 5 frames (ablation) | UST-Hand · HanCo | MPVPE: 5.94 -> 5.82 | Improved |
| increase temporal window from 5 to 7 frames (ablation) | UST-Hand · HanCo | MPVPE: 5.82 -> 5.87 | Hurt |
| increase STPT blocks from 1 to 4 (ablation) | UST-Hand · HanCo | MPVPE: 6.01 -> 5.82 | Improved |
Closest papers by meaning
- SO-HandNet: Self-Organizing Network for 3D Hand Pose Estimation With Semi-Supervised Learning (ICCV 2019 · similarity 0.90)
- Two Heads Are Better than One: Image-Point Cloud Network for Depth-Based 3D Hand Pose Estimation (AAAI 2023 · similarity 0.89)
- Tracking and Reconstructing Hand Object Interactions from Point Cloud Sequences in the Wild (AAAI 2023 · similarity 0.89)
- Leveraging Photometric Consistency Over Time for Sparsely Supervised Hand-Object Reconstruction (CVPR 2020 · similarity 0.88)
- Point-to-Point Regression PointNet for 3D Hand Pose Estimation (ECCV 2018 · similarity 0.88)
- HOISDF: Constraining 3D Hand-Object Pose Estimation with Global Signed Distance Fields (CVPR 2024 · 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.
