Research map / Optimization / Nonconvex and smooth optimization
Hidden Convexity of Wasserstein GANs: Interpretable Generative Models with Closed-Form Solutions
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 ProCoGAN instead of Progressive GDA | Linear generators with quadratic discriminators · CelebA | FID: 194.1 ± 4.5 -> 128.4 ± 0.4 | Helped |
| Use ProCoGAN instead of Progressive GDA | Linear generators with quadratic discriminators · CelebA | training time (seconds): 11696 ± 81 -> 153 ± 3 seconds | Helped |
| Increase β_d^(4), β_d^(5) from (7.2×10^3, 1.0×10^4) to (1.9×10^4, 3.3×10^4) (ablation) | ProCoGAN · CelebA | FID: 128.4 ± 0.4 -> 147.1 ± 2.4 | Hurt |
| Increase discriminator penalty β_d (ablation) | ProCoGAN · CelebA generated faces | image diversity | Hurt |
| Increase discriminator penalty β_d (ablation) | ProCoGAN · CelebA generated faces | image fidelity | Helped |
| Increase generator penalty β_g (ablation) | Progressive GDA · CelebA generated faces | image diversity | Hurt |
| Increase generator penalty β_g (ablation) | Progressive GDA · CelebA generated faces | image noise | Helped |
Closest papers by meaning
- A Convex Duality Framework for GANs (NeurIPS 2018 · similarity 0.94)
- Don't let your Discriminator be fooled (ICLR 2019 · similarity 0.92)
- TRAINING GENERATIVE ADVERSARIAL NETWORKS VIA PRIMAL-DUAL SUBGRADIENT METHODS: A LAGRANGIAN PERSPECTIVE ON GAN (ICLR 2018 · similarity 0.92)
- On the Convergence and Robustness of Training GANs with Regularized Optimal Transport (NeurIPS 2018 · similarity 0.91)
- Improving the Improved Training of Wasserstein GANs: A Consistency Term and Its Dual Effect (ICLR 2018 · similarity 0.91)
- Generative Modeling Using the Sliced Wasserstein Distance (CVPR 2018 · similarity 0.91)
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.