Research map / Image generation / GANs
Wasserstein 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 |
|---|---|---|---|
| Replace JS-divergence GAN objective with Wasserstein-1 (EM) critic objective with weight clipping | GAN training · LSUN-Bedrooms, DCGAN generator | sample quality (qualitative) | No clear effect |
| Replace standard GAN objective with WGAN critic objective | generator without batch normalization and with constant filters · LSUN-Bedrooms, reduced DCGAN generator | sample quality (qualitative) | Helped |
| Replace standard GAN objective with WGAN critic objective | 4-layer ReLU MLP generator · LSUN-Bedrooms, MLP generator | sample quality and mode collapse (qualitative) | Helped |
| Use WGAN EM-distance loss as training curve instead of JS-based loss | GAN training curves · LSUN-Bedrooms, DCGAN/MLP generators | correlation of loss with sample quality | Helped |
| Switch critic optimizer from Adam (beta1>0) to RMSProp | WGAN critic · LSUN-Bedrooms training stability | training stability | Helped |
| Use high learning rates with WGAN critic | WGAN training · LSUN-Bedrooms training stability | training stability | Hurt |
Closest papers by meaning
- Improved Training of Wasserstein GANs (NeurIPS 2017 · similarity 0.93)
- A Two-Step Computation of the Exact GAN Wasserstein Distance (ICML 2018 · similarity 0.93)
- SGD Learns One-Layer Networks in WGANs (ICML 2020 · similarity 0.92)
- Improving the Improved Training of Wasserstein GANs: A Consistency Term and Its Dual Effect (ICLR 2018 · similarity 0.92)
- Bridging the Gap Between f-GANs and Wasserstein GANs (ICML 2020 · similarity 0.92)
- Banach Wasserstein GAN (NeurIPS 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.