English

Improving dermatology classifiers across populations using images generated by large diffusion models

Image and Video Processing 2022-11-28 v1 Computer Vision and Pattern Recognition Machine Learning

Abstract

Dermatological classification algorithms developed without sufficiently diverse training data may generalize poorly across populations. While intentional data collection and annotation offer the best means for improving representation, new computational approaches for generating training data may also aid in mitigating the effects of sampling bias. In this paper, we show that DALL\cdotE 2, a large-scale text-to-image diffusion model, can produce photorealistic images of skin disease across skin types. Using the Fitzpatrick 17k dataset as a benchmark, we demonstrate that augmenting training data with DALL\cdotE 2-generated synthetic images improves classification of skin disease overall and especially for underrepresented groups.

Keywords

Cite

@article{arxiv.2211.13352,
  title  = {Improving dermatology classifiers across populations using images generated by large diffusion models},
  author = {Luke W. Sagers and James A. Diao and Matthew Groh and Pranav Rajpurkar and Adewole S. Adamson and Arjun K. Manrai},
  journal= {arXiv preprint arXiv:2211.13352},
  year   = {2022}
}

Comments

NeurIPS 2022 Workshop on Synthetic Data for Empowering ML Research

R2 v1 2026-06-28T07:10:54.838Z