Summary of Sampling Strategies For Mitigating Bias in Face Synthesis Methods, by Emmanouil Maragkoudakis et al.
Sampling Strategies for Mitigating Bias in Face Synthesis Methods
by Emmanouil Maragkoudakis, Symeon Papadopoulos, Iraklis Varlamis, Christos Diou
First submitted to arxiv on: 18 May 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computer Vision and Pattern Recognition (cs.CV)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper investigates the potential biases introduced by StyleGAN2, a widely used generative model for synthesizing high-fidelity face images. Specifically, the authors focus on two protected attributes: gender and age. They reveal that biases arise in the distribution of randomly sampled images against very young and very old age groups, as well as female faces. The authors propose two sampling strategies to balance the representation of selected attributes in generated face images. These methods aim to increase the number of generated samples from underrepresented classes by sampling on specific lines or spheres of the latent space. Experimental results show a decrease in bias against underrepresented groups and a more uniform distribution of protected features at different image quality levels. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about making sure that computer-generated images of people don’t have unfair biases. The authors look at how well a popular way to create these images, called StyleGAN2, does this job. They find that some types of people (like very young or old people) are underrepresented in the generated images. To fix this, they propose two new ways to generate these images that make sure all kinds of people are represented fairly. The results show that their methods can reduce bias and create more balanced images. |
Keywords
» Artificial intelligence » Generative model » Latent space
