# On Stabilizing Generative Adversarial Training with Noise

arXiv: https://arxiv.org/abs/1906.04612 · HTML: https://mlsanity.com/paper/1906.04612/

## Key facts

- Cited by 25 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 11 Jun 2019.

## Built on by (most cited first, 25 of 25)

- [Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusion](https://mlsanity.com/paper/2506.08009/) (2025)
- [Regularizing Generative Adversarial Networks under Limited Data](https://mlsanity.com/paper/2104.03310/) (2021)
- [Deceive D: Adaptive Pseudo Augmentation for GAN Training with Limited Data](https://mlsanity.com/paper/2111.06849/) (2021)
- [Free Lunch for Domain Adversarial Training: Environment Label Smoothing](https://mlsanity.com/paper/2302.00194/) (2023)
- [Sparse Inducing Points in Deep Gaussian Processes: Enhancing Modeling with Denoising Diffusion Variational Inference](https://mlsanity.com/paper/2407.17033/) (2024)
- [Generative Adversarial Neural Cellular Automata](https://mlsanity.com/paper/2108.04328/) (2021)
- [Feature Statistics Mixing Regularization for Generative Adversarial Networks](https://mlsanity.com/paper/2112.04120/) (2021)
- [Top-Down Deep Clustering with Multi-generator GANs](https://mlsanity.com/paper/2112.03398/) (2021)
- [MP3net: coherent, minute-long music generation from raw audio with a simple convolutional GAN](https://mlsanity.com/paper/2101.04785/) (2021)
- [Adversarial Attacks and Defenses for Social Network Text Processing Applications: Techniques, Challenges and Future Research Directions](https://mlsanity.com/paper/2110.13980/) (2021)
- [Dual-Space Augmented Intrinsic-LoRA for Wind Turbine Segmentation](https://mlsanity.com/paper/2412.20838/) (2024)
- [Endless World: Real-Time 3D-Aware Long Video Generation](https://mlsanity.com/paper/2512.12430/) (2025)
- [UnfairGAN: An Enhanced Generative Adversarial Network for Raindrop Removal from A Single Image](https://mlsanity.com/paper/2110.05523/) (2021)
- [D3A-TS: Denoising-Driven Data Augmentation in Time Series](https://mlsanity.com/paper/2312.05550/) (2023)
- [Generalization of GANs and overparameterized models under Lipschitz continuity](https://mlsanity.com/paper/2104.02388/) (2021)
- [Self Sparse Generative Adversarial Networks](https://mlsanity.com/paper/2101.10556/) (2021)
- [Generative Adversarial Networks via a Composite Annealing of Noise and Diffusion](https://mlsanity.com/paper/2105.00220/) (2021)
- [Dynamic-Pix2Pix: Noise Injected cGAN for Modeling Input and Target Domain Joint Distributions with Limited Training Data](https://mlsanity.com/paper/2211.08570/) (2022)
- [MS3D: A RG Flow-Based Regularization for GAN Training with Limited Data](https://mlsanity.com/paper/2408.11135/) (2024)
- [Advances in Electron Microscopy with Deep Learning](https://mlsanity.com/paper/2101.01178/) (2021)
- [Prb-GAN: A Probabilistic Framework for GAN Modelling](https://mlsanity.com/paper/2107.05241/) (2021)
- [Extracting Expert's Goals by What-if Interpretable Modeling](https://mlsanity.com/paper/2110.15165/) (2021)
- [Spectral Bandwidth Recovery of Optical Coherence Tomography Images using Deep Learning](https://mlsanity.com/paper/2301.00504/) (2023)
- [Scaling-based Data Augmentation for Generative Models and its Theoretical Extension](https://mlsanity.com/paper/2410.20780/) (2024)
- [Style Quantization for Data-Efficient GAN Training](https://mlsanity.com/paper/2503.24282/) (2025)

## Abstract (arXiv)

We present a novel method and analysis to train generative adversarial networks (GAN) in a stable manner. As shown in recent analysis, training is often undermined by the probability distribution of the data being zero on neighborhoods of the data space. We notice that the distributions of real and generated data should match even when they undergo the same filtering. Therefore, to address the limited support problem we propose to train GANs by using different filtered versions of the real and generated data distributions. In this way, filtering does not prevent the exact matching of the data distribution, while helping training by extending the support of both distributions. As filtering we consider adding samples from an arbitrary distribution to the data, which corresponds to a convolution of the data distribution with the arbitrary one. We also propose to learn the generation of these samples so as to challenge the discriminator in the adversarial training. We show that our approach results in a stable and well-behaved training of even the original minimax GAN formulation. Moreover, our technique can be incorporated in most modern GAN formulations and leads to a consistent improvement on several common datasets.