# Mixed batches and symmetric discriminators for GAN training

arXiv: https://arxiv.org/abs/1806.07185 · HTML: https://mlsanity.com/paper/1806.07185/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Jun 2018.

## Built on by (most cited first, 7 of 7)

- [Instance-Conditioned GAN](https://mlsanity.com/paper/2109.05070/) (2021)
- [Semi-supervised Vision Transformers at Scale](https://mlsanity.com/paper/2208.05688/) (2022)
- [Interpolation-based Contrastive Learning for Few-Label Semi-Supervised Learning](https://mlsanity.com/paper/2202.11915/) (2022)
- [Mixed Graph Contrastive Network for Semi-Supervised Node Classification](https://mlsanity.com/paper/2206.02796/) (2022)
- [Randomized Quantization: A Generic Augmentation for Data Agnostic Self-supervised Learning](https://mlsanity.com/paper/2212.08663/) (2022)
- [A deep learning guided memetic framework for graph coloring problems](https://mlsanity.com/paper/2109.05948/) (2021)
- [Measure-conditional Discriminator with Stationary Optimum for GANs and Statistical Distance Surrogates](https://mlsanity.com/paper/2101.06802/) (2021)

## Abstract (arXiv)

Generative adversarial networks (GANs) are pow- erful generative models based on providing feed- back to a generative network via a discriminator network. However, the discriminator usually as- sesses individual samples. This prevents the dis- criminator from accessing global distributional statistics of generated samples, and often leads to mode dropping: the generator models only part of the target distribution. We propose to feed the discriminator with mixed batches of true and fake samples, and train it to predict the ratio of true samples in the batch. The latter score does not depend on the order of samples in a batch. Rather than learning this invariance, we introduce a generic permutation-invariant discriminator ar- chitecture. This architecture is provably a uni- versal approximator of all symmetric functions. Experimentally, our approach reduces mode col- lapse in GANs on two synthetic datasets, and obtains good results on the CIFAR10 and CelebA datasets, both qualitatively and quantitatively.