# Uncovering Bias in Face Generation Models

arXiv: https://arxiv.org/abs/2302.11562 · HTML: https://mlsanity.com/paper/2302.11562/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Placed in Image generation, text-to-image and editing, one of 14 directions in Diffusion models.
- Builds on 21 of its 29 references that are in the map.
- First posted on arXiv 22 Feb 2023.

## Placed in

- Diffusion models / [Image generation, text-to-image and editing](https://mlsanity.com/diffusion/image-generation-text-to-image-and-editing/)

## Builds on (most cited first, 21 of 21)

- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [A Style-Based Generator Architecture for Generative Adversarial Networks](https://mlsanity.com/paper/1812.04948/) (2018)
- [Analyzing and Improving the Image Quality of StyleGAN](https://mlsanity.com/paper/1912.04958/) (2019)
- [LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop](https://mlsanity.com/paper/1506.03365/) (2015)
- [Imagic: Text-Based Real Image Editing with Diffusion Models](https://mlsanity.com/paper/2210.09276/) (2022)
- [Diffusion Models: A Comprehensive Survey of Methods and Applications](https://mlsanity.com/paper/2209.00796/) (2022)
- [GANSpace: Discovering Interpretable GAN Controls](https://mlsanity.com/paper/2004.02546/) (2020)
- [InterFaceGAN: Interpreting the Disentangled Face Representation Learned by GANs](https://mlsanity.com/paper/2005.09635/) (2020)
- [StyleFlow: Attribute-conditioned Exploration of StyleGAN-Generated Images using Conditional Continuous Normalizing Flows](https://mlsanity.com/paper/2008.02401/) (2020)
- [StyleSpace Analysis: Disentangled Controls for StyleGAN Image Generation](https://mlsanity.com/paper/2011.12799/) (2020)
- [FairGAN: Fairness-aware Generative Adversarial Networks](https://mlsanity.com/paper/1805.11202/) (2018)
- [Fair Generative Modeling via Weak Supervision](https://mlsanity.com/paper/1910.12008/) (2019)
- [Image Generators with Conditionally-Independent Pixel Synthesis](https://mlsanity.com/paper/2011.13775/) (2020)
- [Improving the Fairness of Deep Generative Models without Retraining](https://mlsanity.com/paper/2012.04842/) (2020)
- [Inclusive GAN: Improving Data and Minority Coverage in Generative Models](https://mlsanity.com/paper/2004.03355/) (2020)
- [Fairness GAN](https://mlsanity.com/paper/1805.09910/) (2018)
- [GAN-generated Faces Detection: A Survey and New Perspectives](https://mlsanity.com/paper/2202.07145/) (2022)
- [Markedness in Visual Semantic AI](https://mlsanity.com/paper/2205.11378/) (2022)
- [Characterizing Bias in Classifiers using Generative Models](https://mlsanity.com/paper/1906.11891/) (2019)
- [FairStyle: Debiasing StyleGAN2 with Style Channel Manipulations](https://mlsanity.com/paper/2202.06240/) (2022)

## Built on by (most cited first, 9 of 9)

- [Fair Text-to-Image Diffusion via Fair Mapping](https://mlsanity.com/paper/2311.17695/) (2023)
- [Makeup-Guided Facial Privacy Protection via Untrained Neural Network Priors](https://mlsanity.com/paper/2408.12387/) (2024)
- [Limitations of Face Image Generation](https://mlsanity.com/paper/2309.07277/) (2023)
- [A Picture is Worth 500 Labels: A Case Study of Demographic Disparities in Local Machine Learning Models for Instagram and TikTok](https://mlsanity.com/paper/2403.19717/) (2024)
- [CraftGraffiti: Exploring Human Identity with Custom Graffiti Art via Facial-Preserving Diffusion Models](https://mlsanity.com/paper/2508.20640/) (2025)
- [Adv-TGD: Adversarial Text-Guided Diffusion for Face Recognition Impersonation Attacks](https://mlsanity.com/paper/2606.11615/) (2026)
- [FLUID: Training-Free Face De-identification via Latent Identity Substitution](https://mlsanity.com/paper/2511.17005/) (2025)
- [GLEaN: A Text-to-image Bias Detection Approach for Public Comprehension](https://mlsanity.com/paper/2604.09923/) (2026)
- [Learning Late, Guiding Early: Timestep-Decoupled Semantic Guidance for Fair Face Generation](https://mlsanity.com/paper/2608.25862/) (2026)

## Abstract (arXiv)

Recent advancements in GANs and diffusion models have enabled the creation of high-resolution, hyper-realistic images. However, these models may misrepresent certain social groups and present bias. Understanding bias in these models remains an important research question, especially for tasks that support critical decision-making and could affect minorities. The contribution of this work is a novel analysis covering architectures and embedding spaces for fine-grained understanding of bias over three approaches: generators, attribute modifier, and post-processing bias mitigators. This work shows that generators suffer from bias across all social groups with attribute preferences such as between 75%-85% for whiteness and 60%-80% for the female gender (for all trained CelebA models) and low probabilities of generating children and older men. Modifier and mitigators work as post-processor and change the generator performance. For instance, attribute channel perturbation strategies modify the embedding spaces. We quantify the influence of this change on group fairness by measuring the impact on image quality and group features. Specifically, we use the Fr'echet Inception Distance (FID), the Face Matching Error and the Self-Similarity score. For Interfacegan, we analyze one and two attribute channel perturbations and examine the effect on the fairness distribution and the quality of the image. Finally, we analyzed the post-processing bias mitigators, which are the fastest and most computationally efficient way to mitigate bias. We find that these mitigation techniques show similar results on KL divergence and FID score, however, self-similarity scores show a different feature concentration on the new groups of the data distribution. The weaknesses and ongoing challenges described in this work must be considered in the pursuit of creating fair and unbiased face generation models.