Research map / 3D vision / Human pose estimation
Accelerating Dynamic Programs via Nested Benders Decomposition with Application to Multi-Person Pose Estimation
Shaofei Wang, Alexander Ihler, Konrad Kording, Julian Yarkony
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 |
|---|---|---|---|
| replace exact dynamic programming with nested Benders decomposition for pricing | ICG pricing step · MPII-Multiperson validation | total running time: 44x faster, up to 500x | Improved |
| replace dynamic programming with NBD | ICG pricing step · MPII-Multiperson validation | solution quality | No clear effect |
| use NBD solver instead of primal heuristic solver [10] | MPPE model · MPII-Multiperson validation | wrist AP: 68.4 -> 70.1 | Improved |
| use NBD solver instead of primal heuristic solver [10] | MPPE model · MPII-Multiperson validation | ankle AP: 64.3 -> 64.8 | Improved |
| use NBD solver instead of primal heuristic solver [10] | MPPE model · MPII-Multiperson validation | head AP: 93.0 -> 90.6 | Hurt |
| use NBD solver instead of primal heuristic solver [10] | MPPE model · MPII-Multiperson validation | shoulder AP: 88.2 -> 87.3 | Hurt |
| use NBD solver instead of primal heuristic solver [10] | MPPE model · MPII-Multiperson validation | mAP: 77.6 -> 77.6 | No clear effect |
| use NBD solver instead of primal heuristic solver [10] | MPPE model · MPII-Multiperson validation | time per frame: 0.136 -> 1.95 s/frame | Hurt |
| use NBD solver instead of primal heuristic solver [10] | MPPE model · MPII-Multiperson validation | mAP (UBody): 81.9 -> 81.8 | No clear effect |
Closest papers by meaning
- Joint Graph Decomposition & Node Labeling: Problem, Algorithms, Applications (CVPR 2017 · similarity 0.87)
- End-to-End Learning for Graph Decomposition (ICCV 2019 · similarity 0.84)
- Accelerating Globally Optimal Consensus Maximization in Geometric Vision (TPAMI 2024 · similarity 0.84)
- COMPOSE: Hypergraph Cover Optimization for Multi-view 3D Human Pose Estimation (NeurIPS 2026 · similarity 0.84)
- HDG-ODE: A Hierarchical Continuous-Time Model for Human Pose Forecasting (ICCV 2023 · similarity 0.83)
- Multi-target Tracking in Multiple Non-overlapping Cameras Using Fast-Constrained Dominant Sets (IJCV 2019 · similarity 0.83)
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.
