Research map / Image generation / GANs
MSG-GAN: Multi-Scale Gradients for Generative Adversarial Networks
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 |
|---|---|---|---|
| Add multi-scale generator-to-discriminator connections | ProGAN · Oxford Flowers 256x256 | FID: 60.40 -> 28.27 | Helped |
| Add multi-scale generator-to-discriminator connections | StyleGAN · Indian Celebs 256x256 | FID: 61.22 -> 28.44 | Helped |
| Add multi-scale generator-to-discriminator connections | StyleGAN · LSUN Churches 256x256 | FID: 6.58 -> 5.2 | Helped |
| Add multi-scale generator-to-discriminator connections | ProGAN · CelebA-HQ 1024x1024 | FID: 7.79 -> 8.02 | Hurt |
| Add multi-scale generator-to-discriminator connections | StyleGAN · FFHQ 1024x1024 | FID: 4.40 -> 5.8 | Hurt |
| Remove all multi-scale connections (ablation) | MSG-ProGAN · FFHQ 1024x1024 | FID: 8.36 -> 14.20 | Hurt |
| Use coarse-level connections only (ablation) | MSG-ProGAN · FFHQ 1024x1024 | FID: 8.36 -> 10.84 | Hurt |
| Use middle-level connections only (ablation) | MSG-ProGAN · FFHQ 1024x1024 | FID: 8.36 -> 9.17 | Hurt |
| Use fine-level connections only (ablation) | MSG-ProGAN · FFHQ 1024x1024 | FID: 8.36 -> 9.74 | Hurt |
| Use phi_lin_cat instead of phi_simple (ablation) | MSG-ProGAN · FFHQ 1024x1024 | FID: 8.36 -> 11.88 | Hurt |
| Use phi_cat_lin instead of phi_simple (ablation) | MSG-ProGAN · FFHQ 1024x1024 | FID: 8.36 -> 9.63 | Hurt |
| Use phi_cat_lin instead of phi_simple (ablation) | MSG-StyleGAN · FFHQ 1024x1024 | FID: 6.46 -> 5.80 | Helped |
Closest papers by meaning
- RG-GAN: Dynamic Regenerative Pruning for Data-Efficient Generative Adversarial Networks (AAAI 2024 · similarity 0.91)
- Large Scale GAN Training for High Fidelity Natural Image Synthesis (ICLR 2019 · similarity 0.91)
- MCGAN: Enhancing GAN Training with Regression-Based Generator Loss (AAAI 2025 · similarity 0.91)
- MGAN: Training Generative Adversarial Nets with Multiple Generators (ICLR 2018 · similarity 0.91)
- A Large-Scale Study on Regularization and Normalization in GANs (ICML 2019 · similarity 0.91)
- MS$^3$D: A RG Flow-Based Regularization for GAN Training with Limited Data (ICML 2024 · 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.