Research map / Learning theory and algorithms / Optimal transport
Optimal Transport with Symmetry Groups
Jiechao Zhang, Huichun Zhang, Jian Sun, Wei Zeng
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 |
|---|---|---|---|
| use SG-LOT instead of LOT | LOT · Synthetic group O, block ordering, 4,800 points | runtime (sec.): 3.6485 -> 0.1112 | Improved |
| use SG-LOT instead of C-LOT | C-LOT · Synthetic group O, block ordering, 4,800 points | runtime (sec.): 0.3289 -> 0.1112 | Improved |
| use SG-EROT instead of EROT | EROT · Synthetic group O, block ordering, 9,600 points | runtime (sec.): 11.6585 -> 0.6789 | Improved |
| use SG-EROT instead of EROT under random ordering | EROT · Synthetic group O, random ordering, 9,600 points | runtime (sec.): 10.4986 -> 4.1449 | Improved |
| use SG-LOT instead of LOT under random ordering | LOT · Synthetic group O, random ordering, 9,600 points | runtime (sec.): 18.1031 -> 4.1651 | Improved |
| use SG-EROT instead of EROT | EROT · NYU images D2, block ordering, 9,600 points | runtime (sec.): 50.5090 -> 4.1361 | Improved |
| use SG-EROT instead of C-EROT | C-EROT · NYU images D2, block ordering, 9,600 points | runtime (sec.): 15.9426 -> 4.1361 | Improved |
| use SG-EROT instead of EROT under random ordering | EROT · NYU images D2, random ordering, 4,800 points | runtime (sec.): 13.9812 -> 2.1840 | Improved |
| use SG-EROT instead of EROT | EROT · Molecular graph matching | runtime (ms): 0.5850 -> 0.2909 | Improved |
| use SG-EROT instead of EROT on exact symmetry | EROT · Synthetic group O, 4,800 points | objective value: -0.9248 -> -0.9248 | No clear effect |
| use SG-EROT instead of EROT on approximate symmetry | EROT · NYU images D2, block ordering, 4,800 points | objective value: -0.9789 -> -0.9909 | No clear effect |
| add Gaussian noise to the upper-left quadrant (variances 0 to 640) before SG-EROT (ablation) | SG-EROT · NYU images, perturbed symmetry | gap to direct EROT solution | Hurt |
Closest papers by meaning
- Optimal Transport with Cyclic Symmetry (AAAI 2024 · similarity 0.89)
- A Truncated Newton Method for Optimal Transport (ICLR 2025 · similarity 0.86)
- Progressive Entropic Optimal Transport Solvers (NeurIPS 2024 · similarity 0.85)
- A Computational Framework for Solving Wasserstein Lagrangian Flows (ICML 2024 · similarity 0.85)
- Safe and Sparse Newton Method for Entropic-Regularized Optimal Transport (NeurIPS 2024 · similarity 0.85)
- A Provably Convergent and Practical Algorithm for Gromov–Wasserstein Optimal Transport (NeurIPS 2026 · similarity 0.85)
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.
