Summary of Fairdd: Fair Dataset Distillation Via Synchronized Matching, by Qihang Zhou et al.
FairDD: Fair Dataset Distillation via Synchronized Matching
by Qihang Zhou, Shenhao Fang, Shibo He, Wenchao Meng, Jiming Chen
First submitted to arxiv on: 29 Nov 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Machine Learning (cs.LG)
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 addresses the concern of ensuring fair image recognition models when training on condensed datasets. The authors show that previous dataset distillation (DD) methods fail to alleviate bias towards minority groups in original datasets, which worsens in the condensed datasets. To bridge this gap, they propose a novel framework called FairDD, which synchronously matches synthetic datasets to protected attribute-wise groups of original datasets. This allows for balanced generation of all protected attribute groups and regularizes vanilla DDs to favor biased generation towards minority groups while maintaining classification accuracy. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary In simple terms, this paper is about making sure that computer vision models are fair when trained on smaller versions of big datasets. Current methods don’t work well because they make the bias in the original data even worse. To fix this, the authors introduce a new way to condense datasets called FairDD, which ensures that the condensed dataset represents all groups equally and fairly. |
Keywords
» Artificial intelligence » Classification » Distillation
