Research map / Image generation / GANs
Lipschitz Generative Adversarial Nets
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 |
|---|---|---|---|
| Apply a quadratic penalty to the discriminator's Lipschitz constant | GAN objectives (LGAN) · Unsupervised image generation | sample quality | Helped |
| Use the exponential loss objective (ablation) | LGAN · Unsupervised image generation | sample quality | Helped |
| Use the x+sqrt(x^2+1) loss objective (ablation) | LGAN · Unsupervised image generation | sample quality | Helped |
| Vary the loss objective among the tested LGAN losses (ablation) | LGAN · CIFAR-10 and two-dimensional toy data | gradient direction toward real samples | No clear effect |
| Use hinge loss with a suitable penalty coefficient (ablation) | LGAN · Unsupervised image generation | sample quality | Helped |
| Use quadratic loss with a suitable penalty coefficient (ablation) | LGAN · Unsupervised image generation | sample quality | Helped |
| Replace GP or LP with MaxGP (ablation) | Lipschitz-regularized GAN training · Synthetic experiments and training with exponential loss | training convergence | Mixed |
| Penalize the Lipschitz constant | LGAN · GAN theory | existence of a unique Nash equilibrium at matching distributions | Helped |
| Use an unrestricted discriminator instead of a Lipschitz-constrained discriminator | Traditional GAN objectives · Disjoint real and generated supports | informativeness of the generator gradient | Hurt |
Closest papers by meaning
- Adversarial Lipschitz Regularization (ICLR 2020 · similarity 0.93)
- On the regularization of Wasserstein GANs (ICLR 2018 · similarity 0.92)
- Towards Generalized Implementation of Wasserstein Distance in GANs (AAAI 2021 · similarity 0.92)
- Improving the Improved Training of Wasserstein GANs: A Consistency Term and Its Dual Effect (ICLR 2018 · similarity 0.92)
- Improved Training of Wasserstein GANs (NeurIPS 2017 · similarity 0.91)
- Approximation theory for 1-Lipschitz ResNets (NeurIPS 2025 · similarity 0.87)
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.