# Training Generative Adversarial Networks with Binary Neurons by End-to-end Backpropagation

arXiv: https://arxiv.org/abs/1810.04714 · HTML: https://mlsanity.com/paper/1810.04714/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 10 Oct 2018.

## Built on by (most cited first, 2 of 2)

- [DEPAS: De-novo Pathology Semantic Masks using a Generative Model](https://mlsanity.com/paper/2302.06513/) (2023)
- [Momentum Gradient-based Untargeted Attack on Hypergraph Neural Networks](https://mlsanity.com/paper/2310.15656/) (2023)

## Abstract (arXiv)

We propose the BinaryGAN, a novel generative adversarial network (GAN) that uses binary neurons at the output layer of the generator. We employ the sigmoid-adjusted straight-through estimators to estimate the gradients for the binary neurons and train the whole network by end-to-end backpropogation. The proposed model is able to directly generate binary-valued predictions at test time. We implement such a model to generate binarized MNIST digits and experimentally compare the performance for different types of binary neurons, GAN objectives and network architectures. Although the results are still preliminary, we show that it is possible to train a GAN that has binary neurons and that the use of gradient estimators can be a promising direction for modeling discrete distributions with GANs. For reproducibility, the source code is available at https://github.com/salu133445/binarygan .