Research map / Learning theory and algorithms / Optimal transport
Learning Wasserstein Embeddings
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 |
|---|---|---|---|
| Replace repeated optimal-transport distance computation with learned Euclidean embeddings | Wasserstein distance computation on MNIST · MNIST test | Speedup: x18 and x1000 on CPU; x1000 and x500 000 on GPU | Helped |
| Train the embedding on the same dataset being evaluated rather than a different dataset | DWE · Google Doodle Cat | MSE: 5.025 -> 1.195 | Helped |
| Train the embedding on the same dataset being evaluated rather than a different dataset | DWE · Google Doodle Crab | MSE: 5.532 -> 0.854 | Helped |
| Train the embedding on the same dataset being evaluated rather than a different dataset | DWE · Google Doodle Faces | MSE: 5.025 -> 3.158 | Helped |
| Train the embedding on the same dataset being evaluated rather than a different dataset | DWE · MNIST | MSE: 50.527 -> 0.405 | Helped |
| Use an embedding trained on another doodle dataset | DWE · Google Doodle datasets | MSE: Cat: 1.195 vs 1.718, 2.07; Crab: 0.854 vs 1.718, 3.158; Faces: 3.158 vs 2.07, 3.158 | Mixed |
| Add a decoder reconstruction loss to the embedding objective | DWE · Image datasets | Generalization performance | Helped |
Closest papers by meaning
- Wasserstein of Wasserstein Loss for Learning Generative Models (ICML 2019 · similarity 0.91)
- Fourier Sliced-Wasserstein Embedding for Multisets and Measures (ICLR 2025 · similarity 0.88)
- Tree-Wasserstein Distance for High Dimensional Data with a Latent Feature Hierarchy (ICLR 2025 · similarity 0.88)
- Statistical and Computational Guarantees of Kernel Max-Sliced Wasserstein Distances (ICML 2025 · similarity 0.87)
- Wasserstein Wormhole: Scalable Optimal Transport Distance with Transformer (ICML 2024 · similarity 0.86)
- A New Robust Partial p-Wasserstein-Based Metric for Comparing Distributions (ICML 2024 · similarity 0.86)
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.