Research map / Image generation / GANs
VEEGAN: Reducing Mode Collapse in GANs using Implicit Variational Learning
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 |
|---|---|---|---|
| Use VEEGAN instead of Unrolled GAN | 2D Ring · 2D Ring | modes captured: 7.6 -> 8 | Helped |
| Use VEEGAN instead of Unrolled GAN | 2D Ring · 2D Ring | high-quality samples (%): 35.6 -> 52.9 | Helped |
| Use VEEGAN instead of Unrolled GAN | 2D Grid · 2D Grid | modes captured: 23.6 -> 24.6 | Helped |
| Use VEEGAN instead of Unrolled GAN | 2D Grid · 2D Grid | high-quality samples (%): 16 -> 40 | Helped |
| Use VEEGAN instead of ALI | 1200D Synthetic · 1200D Synthetic | modes captured: 3 -> 5.5 | Helped |
| Use VEEGAN instead of ALI | 1200D Synthetic · 1200D Synthetic | high-quality samples (%): 5.4 -> 28.29 | Helped |
| Use VEEGAN instead of DCGAN | Stacked-MNIST · Stacked-MNIST | modes captured: 99 -> 150 | Helped |
| Use VEEGAN instead of DCGAN | Stacked-MNIST · Stacked-MNIST | KL divergence: 3.4 -> 2.95 | Helped |
| Use VEEGAN instead of ALI | CIFAR-10 · CIFAR-10 | IvOM: 0.0067 ± 0.004 -> 0.0068 ± 0.0001 | No clear effect |
| Replace the noise autoencoder loss with a data autoencoder loss (ablation) | VEEGAN on CIFAR-10 · CIFAR-10 | sample quality | Hurt |
| Pretrain the reconstructor on real data (ablation) | VEEGAN · Not specified | performance | Helped |
Closest papers by meaning
- Dual Discriminator Generative Adversarial Nets (NeurIPS 2017 · similarity 0.88)
- Re-GAN: Data-Efficient GANs Training via Architectural Reconfiguration (CVPR 2023 · similarity 0.88)
- A Geometric Analysis of Deep Generative Image Models and Its Applications (ICLR 2021 · similarity 0.88)
- MGAN: Training Generative Adversarial Nets with Multiple Generators (ICLR 2018 · similarity 0.88)
- MCGAN: Enhancing GAN Training with Regression-Based Generator Loss (AAAI 2025 · similarity 0.87)
- The GAN is dead; long live the GAN! A Modern GAN Baseline (NeurIPS 2024 · similarity 0.87)
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.