Research map / 3D vision / Human pose estimation
Structural Triangulation: A Closed-Form Solution to Constrained 3D Human Pose Estimation
Zhuo Chen, Xu Zhao, Xiaoyue Wan
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 Step Constraint Algorithm (ablation) | Structural Triangulation · Human3.6M | MPJPE-re: 20.22 -> 19.35 | Improved |
| add Step Constraint Algorithm (ablation) | Structural Triangulation · Human3.6M | MPJPE-ab: 17.98 -> 17.78 | Improved |
| add structural triangulation with GT bone lengths | SVD/algebraic triangulation baseline · Human3.6M | MPJPE-re: 20.35 -> 19.35 | Improved |
| add structural triangulation with GT bone lengths | algebraic triangulation baseline · Total Capture | MPJPE-re: 63.3 -> 59.6 | Improved |
| use estimated bone lengths from T-pose instead of GT | Structural Triangulation · Total Capture | MPJPE-re: 63.3 -> 66.9 | Hurt |
| use estimated bone lengths from T-pose instead of GT (ablation) | Structural Triangulation · Human3.6M | MPJPE-re: 19.35 -> 20.86 | Hurt |
| add structural triangulation without SCA, estimated bone lengths (ablation) | algebraic baseline · Human3.6M S9 | MPBLE: 12.2 -> 7.93 | Improved |
| add SCA to ST with estimated bone lengths (ablation) | Structural Triangulation · Human3.6M S9 | MBLS: 0.350 -> 0.122 | Improved |
| add SCA to ST with GT bone lengths (ablation) | Structural Triangulation · Human3.6M S9 | MPBLE: 1.31 -> 0.129 | Improved |
| increase SCA steps to N=9 (ablation) | Structural Triangulation + SCA · Human3.6M | MPJPE-re: 19.33 | Improved |
| increase SCA steps to N>=10 (ablation) | Structural Triangulation + SCA · Human3.6M | MPJPE-re | Hurt |
| add structural triangulation to synthesized noisy 2D estimates | SVD triangulation · Synthesized 2D (sigma 2-20 px, 2-10 cameras) | absolute MPJPE: more than 82% | Improved |
Closest papers by meaning
- Learnable Triangulation of Human Pose (ICCV 2019 · similarity 0.89)
- Robust 3D Human Pose Estimation from Single Images or Video Sequences (TPAMI 2018 · similarity 0.89)
- 3D Pictorial Structures Revisited: Multiple Human Pose Estimation (TPAMI 2016 · similarity 0.89)
- Keep It SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Image (ECCV 2016 · similarity 0.88)
- Probabilistic Triangulation for Uncalibrated Multi-View 3D Human Pose Estimation (ICCV 2023 · similarity 0.88)
- Toward Approaches to Scalability in 3D Human Pose Estimation (NeurIPS 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.
