# On Noise Injection in Generative Adversarial Networks

arXiv: https://arxiv.org/abs/2006.05891 · HTML: https://mlsanity.com/paper/2006.05891/

## Key facts

- Cited by 19 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 10 Jun 2020.

## Built on by (most cited first, 19 of 19)

- [Generalizable Synthetic Image Detection via Language-guided Contrastive Learning](https://mlsanity.com/paper/2305.13800/) (2023)
- [Rank Diminishing in Deep Neural Networks](https://mlsanity.com/paper/2206.06072/) (2022)
- [Diffusion Model-Augmented Behavioral Cloning](https://mlsanity.com/paper/2302.13335/) (2023)
- [Beyond Augmentation: Empowering Model Robustness under Extreme Capture Environments](https://mlsanity.com/paper/2407.13640/) (2024)
- [HairCLIPv2: Unifying Hair Editing via Proxy Feature Blending](https://mlsanity.com/paper/2310.10651/) (2023)
- [Weakly Supervised High-Fidelity Clothing Model Generation](https://mlsanity.com/paper/2112.07200/) (2021)
- [Makeup Prior Models for 3D Facial Makeup Estimation and Applications](https://mlsanity.com/paper/2403.17761/) (2024)
- [Learning Multiple Probabilistic Degradation Generators for Unsupervised Real World Image Super Resolution](https://mlsanity.com/paper/2201.10747/) (2022)
- [Precise and Generalized Robustness Certification for Neural Networks](https://mlsanity.com/paper/2306.06747/) (2023)
- [Out-of-domain GAN inversion via Invertibility Decomposition for Photo-Realistic Human Face Manipulation](https://mlsanity.com/paper/2212.09262/) (2022)
- [Controlled GAN-Based Creature Synthesis via a Challenging Game Art Dataset -- Addressing the Noise-Latent Trade-Off](https://mlsanity.com/paper/2108.08922/) (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)
- [Finding mixed-strategy equilibria of continuous-action games without gradients using randomized policy networks](https://mlsanity.com/paper/2211.15936/) (2022)
- [Efficiency without Compromise: CLIP-aided Text-to-Image GANs with Increased Diversity](https://mlsanity.com/paper/2506.01493/) (2025)
- [Maximum Likelihood on the Joint (Data, Condition) Distribution for Solving Ill-Posed Problems with Conditional Flow Models](https://mlsanity.com/paper/2208.11782/) (2022)
- [Defensive ML: Defending Architectural Side-channels with Adversarial Obfuscation](https://mlsanity.com/paper/2302.01474/) (2023)
- [Noise Dimension of GAN: An Image Compression Perspective](https://mlsanity.com/paper/2403.09196/) (2024)
- [Scaling-based Data Augmentation for Generative Models and its Theoretical Extension](https://mlsanity.com/paper/2410.20780/) (2024)

## Abstract (arXiv)

Noise injection has been proved to be one of the key technique advances in generating high-fidelity images. Despite its successful usage in GANs, the mechanism of its validity is still unclear. In this paper, we propose a geometric framework to theoretically analyze the role of noise injection in GANs. Based on Riemannian geometry, we successfully model the noise injection framework as fuzzy equivalence on the geodesic normal coordinates. Guided by our theories, we find that the existing method is incomplete and a new strategy for noise injection is devised. Experiments on image generation and GAN inversion demonstrate the superiority of our method.