Research map / Learning theory and algorithms / Optimal transport
Asymptotic Guarantees for Learning Generative Models with the Sliced-Wasserstein Distance
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 the Sliced-Wasserstein distance for minimum-distance estimation | Generative models with intractable likelihoods · General Wasserstein spaces | weak convergence | Helped |
| Use minimum Sliced-Wasserstein estimation | Parametric generative models · Consistency as the number of observations increases | estimator consistency | Helped |
| Use minimum expected Sliced-Wasserstein estimation | Parametric generative models · Consistency as observations and generated samples increase | estimator consistency | Helped |
| Increase the number of generated samples used by MESWE | MESWE · Fixed observed dataset | convergence to MSWE | Helped |
| Use the order-1 Sliced-Wasserstein distance for MSWE | MSWE · Well-specified models in any finite dimension | convergence rate: √n | Helped |
| Replace multivariate Wasserstein estimation with MESWE | Elliptically contoured stable distributions · Parameter inference at higher dimensions | computational time | Helped |
| Use MESWE with the SW generator across training-set sizes and generated-sample counts | Sliced-Wasserstein Generator · MNIST | training and test loss | Mixed |
Closest papers by meaning
- Asymptotic Guarantees for Generative Modeling Based on the Smooth Wasserstein Distance (NeurIPS 2020 · similarity 0.93)
- Distributional Sliced-Wasserstein and Applications to Generative Modeling (ICLR 2021 · similarity 0.92)
- Generative Modeling Using the Sliced Wasserstein Distance (CVPR 2018 · similarity 0.91)
- Sliced Wasserstein Estimation with Control Variates (ICLR 2024 · similarity 0.90)
- Towards Understanding Gradient Dynamics of the Sliced-Wasserstein Distance via Critical Point Analysis (ICML 2025 · similarity 0.89)
- Statistically Optimal Generative Modeling with Maximum Deviation from the Empirical Distribution (ICML 2024 · similarity 0.89)
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.