Research map / Image generation / GANs
Normalized Diversification
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 normalized diversification to GAN+SN | GAN+SN · CIFAR-10 | FID: 23.7 -> 22.9 | Helped |
| Add normalized diversification to GAN+SN+GP | GAN+SN+GP · CIFAR-10 | FID: 22.9 -> 22.0 | Helped |
| Add normalized diversification to GAN+SN | GAN+SN · CelebA | FID: 10.5 -> 10.2 | Helped |
| Add normalized diversification to WGAN+SN+GP | WGAN+SN+GP · CelebA | FID: 9.2 -> 9.0 | Helped |
| Add normalized diversification to GAN-CLS | GAN-CLS · Oxford-102 Flower test | diversity: 43.7±16.8 -> 58.1±18.5 | Helped |
| Add normalized diversification and remove the conditional latent regression branch | BicycleGAN · Facade image generation | FID: 83.0 -> 76.3 | Helped |
| Add normalized diversification and remove the conditional latent regression branch | BicycleGAN · Facade image generation | LPIPS: 0.136 ± 0.049 -> 0.188 ± 0.048 | Helped |
| Add normalized diversity loss (ablation) | Conditional hand pose model · GANerated hands | visible joint accuracy: 91.7 -> 92.8 | Helped |
| Use higher-dimensional latent features (ablation) | Conditional hand pose model · GANerated hands | visible joint accuracy: 92.8 -> 93.2 | Helped |
| Use normalized diversification rather than deterministic regression | Hand pose estimation model · GANerated hands | visible joint accuracy: 91.6 -> 92.8 | Helped |
| Use normalized diversification for multimodal pose prediction | Hand pose estimation model · Stereo Tracking Benchmark | 3D PCK AUC: 0.955 -> 0.980 | Helped |
| Use normalized diversification instead of BourGAN | Synthetic 2D Gaussian ring generator · 2D Ring | fail (%): 0.1 -> 10.3 | Hurt |
Closest papers by meaning
- Diverse Rare Sample Generation with Pretrained GANs (AAAI 2025 · similarity 0.90)
- Diversity-Sensitive Conditional Generative Adversarial Networks (ICLR 2019 · similarity 0.89)
- Diverse Image Generation via Self-Conditioned GANs (CVPR 2020 · similarity 0.89)
- UniGAN: Reducing Mode Collapse in GANs using a Uniform Generator (NeurIPS 2022 · similarity 0.89)
- CHAIN: Enhancing Generalization in Data-Efficient GANs via lipsCHitz continuity constrAIned Normalization (CVPR 2024 · similarity 0.86)
- MS$^3$D: A RG Flow-Based Regularization for GAN Training with Limited Data (ICML 2024 · similarity 0.86)
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.