← Back to Discoveries | bioRxiv 2026-02-12 synthesized

Machine Learning Predicts Drought Tolerance from Leaf Spectral Signatures

Osei-Bonsu K, Rivera M, Huang Y

Summary

bioRxiv

AI can predict drought tolerance from leaf color alone with 89% accuracy, enabling rapid screening of crops without damaging plants.

chevron_right Technical Details

Key Findings

1

89% accuracy across 15 species

2

Near-infrared spectral shifts as key features

3

Non-destructive rapid phenotyping

description

Original Abstract

A convolutional neural network trained on hyperspectral leaf images predicts drought tolerance with 89% accuracy across 15 crop species. The model identifies subtle spectral shifts in the 1400-1900nm range as primary predictive features, enabling rapid phenotyping without destructive sampling.

hub

This connects to 10 other discoveries — 3 species, 4 topics, 3 related articles

Species Mentioned

Wheat
eco Wheat

Wheat is a group of wild and domesticated grasses of the genus Triticum. As cereals, they are cultivated for their grains, which are staple foods around the world. Well-known wheat species and hybrids include the most widely grown common wheat, spelt, durum, emmer, einkorn, and Khorasan or Kamut....

open_in_new Wikipedia