Research map / Learning theory and algorithms / Prediction and decision losses
Efficient and Consistent Adversarial Bipartite Matching
Rizal Fathony, Sima Behpour, Xinhua Zhang, Brian Ziebart
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 |
|---|---|---|---|
| adversarial bipartite matching (marginal) instead of SSVM | Bipartite matching · Bahnhof/Sunnyday | accuracy: 0.729 -> 0.754 | Improved |
| adversarial matching (double oracle) instead of SSVM | Bipartite matching · Pedcross2/Bahnhof | accuracy: 0.731 -> 0.760 | Improved |
| adversarial matching instead of SSVM | Bipartite matching · Campus/Stadtmitte (TUD) | accuracy: 0.662 -> 0.662 | No clear effect |
| adversarial matching instead of SSVM | Bipartite matching · Stadtmitte/Campus (TUD) | accuracy: 0.660 -> 0.672 | Improved |
| double-oracle prediction instead of marginal prediction (ablation) | Adversarial matching · ETH/TUD pairs | accuracy: 0.718 -> 0.714 | No clear effect |
| marginal prediction vs double oracle on Pedcross2/Bahnhof (ablation) | Adversarial matching · Pedcross2/Bahnhof | accuracy: 0.760 -> 0.763 | Improved |
| increase number of elements from 12 to 34 (ablation) | Adversarial matching (marginal) · Training time | time (s): 1.96 -> 9.79 | Hurt |
| SSVM instead of marginal adversarial for training speed | SSVM · Training time, 34 elements | time (s): 9.79 -> 0.31 | Improved |
| double oracle instead of marginal training (ablation) | Adversarial matching · Training time | time | Hurt |
| small number of augmenting permutations in double oracle (ablation) | Adversarial matching DO · Camp/Stadtmitte to Sunnyday/Pedcross2 | number of permutations: 5.8 -> 14.4 | Improved |
Closest papers by meaning
- Active Learning for Probabilistic Structured Prediction of Cuts and Matchings (ICML 2019 · similarity 0.87)
- Adversarial Structure Matching for Structured Prediction Tasks (CVPR 2019 · similarity 0.85)
- Expressive Score-Based Priors for Distribution Matching with Geometry-Preserving Regularization (ICML 2025 · similarity 0.83)
- Online bipartite matching with imperfect advice (ICML 2024 · similarity 0.83)
- SMaRt: Improving GANs with Score Matching Regularity (ICML 2024 · similarity 0.83)
- Diffusion Model for Dense Matching (ICLR 2024 · 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.
