Research map / Image generation / GANs
Training Generative Adversarial Networks in One Stage
Check 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 two-stage training with one-stage synchronous updates | Symmetric DCGAN · CelebA | FID: 23.78±0.12 -> 22.66±0.09 | Helped |
| Replace two-stage training with one-stage synchronous updates | Symmetric DCGAN · LSUN Churches | FID: 46.09±0.25 -> 38.99±0.21 | Helped |
| Replace two-stage training with one-stage synchronous updates | Symmetric DCGAN · FFHQ | FID: 40.56±0.32 -> 29.38±0.24 | Helped |
| Replace two-stage training with one-stage synchronous updates | Symmetric WGAN · FFHQ | FID: 35.66±0.13 -> 39.77±0.14 | Hurt |
| Replace two-stage training with one-stage updates and gradient decomposition | Asymmetric GAN variants · CelebA, LSUN Churches, and FFHQ | FID | Helped |
| Replace two-stage training with one-stage updates and gradient decomposition | Asymmetric GeoGAN, RelGAN, and BGAN · CelebA, LSUN Churches, and FFHQ | FID | Mixed |
| Replace two-stage training with one-stage training | Adversarial knowledge distillation · CIFAR10 | accuracy: DFAD: 0.933; Ours(one): 0.934 | No clear effect |
| Replace two-stage training with one-stage training | Adversarial knowledge distillation · CIFAR100 | accuracy: DFAD: 0.677; Ours(one): 0.700 | Helped |
| Use one-stage symmetric DCGAN training | DCGAN-sym · CelebA and LSUN Churches | training speedup: 458.7/267.5 ≈ 1.7 | Helped |
| Use one-stage asymmetric DCGAN training | DCGAN-asym · CelebA and LSUN Churches | training speedup: 286.1/180.5 ≈ 1.6 | Helped |
| Use one-stage asymmetric LSGAN training | LSGAN-asym · All evaluated datasets | training speedup: 286.4/186.0 ≈ 1.5 | Helped |
Closest papers by meaning
- Training Generative Adversarial Networks with Limited Data (NeurIPS 2020 · similarity 0.92)
- MGAN: Training Generative Adversarial Nets with Multiple Generators (ICLR 2018 · similarity 0.91)
- An Online Learning Approach to Generative Adversarial Networks (ICLR 2018 · similarity 0.91)
- GLeaD: Improving GANs With a Generator-Leading Task (CVPR 2023 · similarity 0.91)
- DO-GAN: A Double Oracle Framework for Generative Adversarial Networks (CVPR 2022 · similarity 0.90)
- MCGAN: Enhancing GAN Training with Regression-Based Generator Loss (AAAI 2025 · similarity 0.89)
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.