Research map / 3D vision / Human pose estimation
Generalizable Human Pose Triangulation
Kristijan Bartol, David Bojanić, Tomislav Petković, Tomislav Pribanić
Check 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 stochastic generalizable triangulation instead of Iskakov et al.'s volumetric approach | Human3.6M after training on CMU3 · CMU Panoptic Studio to Human3.6M | MPJPE (mm): 34.0\,mm -> 31.0\,mm; 8.8% | Improved |
| Use stochastic generalizable triangulation instead of RANSAC | CMU Panoptic Studio novel camera views · CMU Panoptic Studio | MPJPE (mm): 39.5 -> 25.4 | Improved |
| Use stochastic generalizable triangulation instead of VoxelPose | CMU Panoptic Studio novel camera views · CMU Panoptic Studio | MPJPE (mm): 25.5 -> 25.4 | Improved |
| Use stochastic generalizable triangulation instead of Epipolar transformers | Human3.6M, Protocol 1, abs. positions · Human3.6M | MPJPE (mm): 26.9 -> 29.1 | Hurt |
| Use weighted hypothesis averaging rather than an unweighted average (ablation) | Generalizable Human Pose Triangulation · Human3.6M | MPJPE (mm): 31.2 -> 29.1 | Improved |
| Use weighted hypothesis averaging rather than selecting the most probable hypothesis (ablation) | Generalizable Human Pose Triangulation · Human3.6M | MPJPE (mm): 41.3 -> 29.1 | Improved |
| Use weighted hypothesis averaging rather than stochastic hypothesis selection (ablation) | Generalizable Human Pose Triangulation · Human3.6M | MPJPE (mm): 41.3 -> 29.1 | Improved |
| Use weighted hypothesis averaging rather than random hypothesis selection (ablation) | Generalizable Human Pose Triangulation · CMU Panoptic Studio | MPJPE (mm): 26.1 -> 24.9 | Improved |
| Use weighted hypothesis averaging rather than selecting the most probable hypothesis (ablation) | Generalizable Human Pose Triangulation · CMU Panoptic Studio | MPJPE (mm): 25.0 -> 24.9 | No clear effect |
| Estimate relative camera rotation while keeping translation known (ablation) | Generalizable Human Pose Triangulation · Human3.6M, base camera arrangement | MPJPE (mm): 29.1\,mm -> 29.4\,mm | Hurt |
| Estimate relative camera translation while keeping rotation known (ablation) | Generalizable Human Pose Triangulation · Human3.6M, base camera arrangement | MPJPE (mm): 29.1\,mm -> 36.7\,mm | Hurt |
| Estimate both relative camera rotation and translation (ablation) | Generalizable Human Pose Triangulation · CMU3 to Human3.6M | MPJPE (mm): 31.0\,mm -> 44.5\,mm | Hurt |
Closest papers by meaning
- Probabilistic Triangulation for Uncalibrated Multi-View 3D Human Pose Estimation (ICCV 2023 · similarity 0.92)
- Learnable Triangulation of Human Pose (ICCV 2019 · similarity 0.91)
- MV-SSM: Multi-View State Space Modeling for 3D Human Pose Estimation (CVPR 2025 · similarity 0.89)
- 3D Pictorial Structures Revisited: Multiple Human Pose Estimation (TPAMI 2016 · similarity 0.89)
- VirtualPose: Learning Generalizable 3D Human Pose Models from Virtual Data (ECCV 2022 · similarity 0.88)
- 3D-Aware Hypothesis & Verification for Generalizable Relative Object Pose Estimation (ICLR 2024 · similarity 0.88)
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.
