Synthetic plant disease image generation to improve segmentation tasks in low-resource settings
Elliott Cooper, Zane Hartley, Andrew French · 2026
Read Synthetic plant disease image generation to improve segmentation tasks in low-resource settings by Elliott Cooper, Zane Hartley, Andrew French (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13194994.txt.
Description
Elliott Cooper, Zane Hartley, and Andrew French build a pipeline that creates fake plant disease images to help train image models, in work published in 2026. Marking diseased leaf regions by hand takes large labelled datasets, which are scarce for rare diseases and poor regions. Synthetic data may fill the gap.
Their pipeline first builds three-dimensional leaf models in a graphics package, which gives exact ground-truth masks for free. A diffusion image model, fine-tuned on disease photos, then paints realistic disease onto the leaves while control tools keep the leaf shape and lesion placement right. They tested the method on apple leaf diseases with a deliberately small real dataset. Mixing pipeline images with the limited real data improved disease segmentation scores steadily.
Catalog record
| CREATOR | Elliott Cooper, Zane Hartley, Andrew French |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |