Research map / Image generation / GANs
Metropolis-Hastings Generative Adversarial Networks
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 |
|---|---|---|---|
| Wrap the DCGAN generator with calibrated Metropolis-Hastings sampling | DCGAN · CIFAR-10 | Inception score: 2.8789 -> 3.379(66) | Helped |
| Wrap the DCGAN generator with calibrated Metropolis-Hastings sampling | DCGAN · CelebA | Inception score: 2.3317 -> 3.106(64) | Helped |
| Wrap the WGAN generator with calibrated Metropolis-Hastings sampling | WGAN · CIFAR-10 | Inception score: 3.0734 -> 3.305(83) | Helped |
| Wrap the WGAN generator with calibrated Metropolis-Hastings sampling | WGAN · CelebA | Inception score: 2.7876 -> 2.889(89) | Helped |
| Calibrate discriminator probabilities before Metropolis-Hastings sampling (ablation) | MH-GAN with DCGAN · CIFAR-10 | Inception score: 3.113(69) -> 3.379(66) | Helped |
| Calibrate discriminator probabilities before Metropolis-Hastings sampling (ablation) | MH-GAN with DCGAN · CelebA | Inception score: 2.682(50) -> 3.106(64) | Helped |
| Calibrate discriminator probabilities before Metropolis-Hastings sampling (ablation) | MH-GAN with WGAN · CIFAR-10 | Inception score | Helped |
| Calibrate discriminator probabilities before Metropolis-Hastings sampling (ablation) | MH-GAN with WGAN · CelebA | Inception score | Helped |
| Use calibrated Metropolis-Hastings sampling instead of calibrated discriminator rejection sampling | DCGAN · CIFAR-10 | Inception score: 3.073(80) -> 3.379(66) | Helped |
| Use calibrated Metropolis-Hastings sampling instead of calibrated discriminator rejection sampling | DCGAN · CelebA | Inception score: 2.869(67) -> 3.106(64) | Helped |
| Use calibrated Metropolis-Hastings sampling instead of calibrated discriminator rejection sampling | WGAN · CIFAR-10 | Inception score: 3.137(64) -> 3.305(83) | Helped |
| Use calibrated Metropolis-Hastings sampling instead of calibrated discriminator rejection sampling | WGAN · CelebA | Inception score: 2.861(66) -> 2.889(89) | Helped |
Closest papers by meaning
- Bayesian GAN (NeurIPS 2017 · similarity 0.91)
- MCGAN: Enhancing GAN Training with Regression-Based Generator Loss (AAAI 2025 · similarity 0.90)
- Variational Annealing of GANs: A Langevin Perspective (ICML 2019 · similarity 0.90)
- MSG-GAN: Multi-Scale Gradients for Generative Adversarial Networks (CVPR 2020 · similarity 0.88)
- ProbGAN: Towards Probabilistic GAN with Theoretical Guarantees (ICLR 2019 · similarity 0.88)
- Statistically Optimal Generative Modeling with Maximum Deviation from the Empirical Distribution (ICML 2024 · 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.