Research map / Image generation / GANs
Chi-square Generative Adversarial Network
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 |
|---|---|---|---|
| Replace GAN/WGAN-GP-style objectives with χ2-GAN (least-squares critic, decoupled updates) on the DCGAN architecture | CIFAR-10 unsupervised generation · CIFAR-10 unsupervised | Inception score: 7.47 vs 6.56 for WGAN-GP (DCGAN) | Helped |
| χ2-GAN objective with ResNet architecture | CIFAR-10 unsupervised · CIFAR-10 unsupervised | Inception score: 7.88 vs 7.86 WGAN-GP ResNet, 7.90 Fisher-GAN ResNet | No clear effect |
| χ2-GAN objective on MNIST | MNIST · MNIST | AIS log p(x): -78.85 vs -79.47 for DCGAN | Helped |
| χ2-GAN objective on ImageNet 64x64 with DCGAN architecture | ImageNet · ImageNet 64x64 | Inception score: 11.34 vs 11.14 ASVAE and 7.65 PixelCNN++ | Helped |
| χ2-GAN supervised objective with ResNet | CIFAR-10 supervised · CIFAR-10 supervised | Inception score: 8.44 vs 8.42 WGAN-GP ResNet but below SGAN 8.59 | Mixed |
| Importance resampling with critic-derived weights on generator samples | face generation (CelebA) and other image models · CelebA | qualitative sample quality: no numeric values reported in main text | Helped |
Closest papers by meaning
- Mixed batches and symmetric discriminators for GAN training (ICML 2018 · similarity 0.91)
- Learning GANs and Ensembles Using Discrepancy (NeurIPS 2019 · similarity 0.90)
- Improving Generalization and Stability of Generative Adversarial Networks (ICLR 2019 · similarity 0.89)
- Dual Discriminator Generative Adversarial Nets (NeurIPS 2017 · similarity 0.89)
- MGAN: Training Generative Adversarial Nets with Multiple Generators (ICLR 2018 · similarity 0.89)
- An Online Learning Approach to Generative Adversarial Networks (ICLR 2018 · 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.