Research map / Image generation / GANs
Improving GAN with Neighbors Embedding and Gradient Matching
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 |
|---|---|---|---|
| add gradient matching to the GAN generator objective (ablation) | standard GAN · CIFAR-10 | FID: 38.10 -> 31.50 | Helped |
| add neighbor embedding to the GAN generator objective (ablation) | standard GAN · CIFAR-10 | FID: 38.10 -> 38.10 | Helped |
| combine gradient matching and neighbor embedding (ablation) | standard GAN · CIFAR-10 | FID: 31.50 -> 26.05 | Helped |
| add neighbor embedding to the original model (ablation) | GN-GAN original model · CIFAR-10 | FID | Helped |
| add gradient matching to the original model (ablation) | GN-GAN original model · CIFAR-10 | FID | Helped |
| combine neighbor embedding and gradient matching (ablation) | GN-GAN original model · CIFAR-10 | FID | Helped |
| use gradient matching for the generator | GAN training · 1D synthetic, three Gaussian modes | mode recovery | Helped |
| use gradient matching for the generator | GAN training · 2D synthetic, 25 Gaussian modes | registered modes | Helped |
| use the combined GN-GAN method | GAN training · CIFAR-10 | FID: 22.95 -> 21.70 | Helped |
| use the combined GN-GAN method | GAN training · STL-10 | FID: 36.19 -> 30.80 | Helped |
Closest papers by meaning
- SMaRt: Improving GANs with Score Matching Regularity (ICML 2024 · similarity 0.91)
- A Large-Scale Study on Regularization and Normalization in GANs (ICML 2019 · similarity 0.91)
- Stabilizing Training of Generative Adversarial Networks through Regularization (NeurIPS 2017 · similarity 0.91)
- Spider GAN: Leveraging Friendly Neighbors To Accelerate GAN Training (CVPR 2023 · similarity 0.90)
- MSG-GAN: Multi-Scale Gradients for Generative Adversarial Networks (CVPR 2020 · 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.