Cross-crop generalization in multi-species leaf disease recognition
4 affiliations
- 1School of Computer and Artificial Intelligence, Research Center of Smart City and Big Data Engineering of Henan Province, Research Center of Intelligent Campus Application Engineering of Henan Province, Henan University of Urban Construction
- 2Binary Graduate School, Binary University of Management & Entrepreneurship
- 3Department of Business Administration, Stamford University Bangladesh
- 4School of Computer and Artificial Intelligence, Henan University of Urban Construction
Abstract
Introduction: Plant diseases are major biotic stress factors whose visual symptoms may be conserved across taxonomic families, though the extent remains unquantified. Cross-crop leaf disease recognition requires architectures that generalize to species absent from training, but existing methods lack a theoretically grounded criterion for ensuring cross-species transferability.
Methods: This study proposes functional-domain separation as such a criterion, derived from the disentangled representation learning hypothesis, and implements it as the leaf-disease composite feature interaction network (LCFIN), a configurable multi-branch feature enhancement framework. LCFIN assigns a two-level Squeeze-and-Excitation mechanism to a channel calibration domain and an MLP-Mixer head to a spatial interaction domain on an EfficientNet-B3 backbone, with dual-weighted sampling addressing cross-level class imbalance. It is trained for main fruit and vegetable leaf disease identification (MFVLDI) on three families (Rosaceae, Vitaceae, Solanaceae) and evaluated under a frozen-representation linear-probe protocol on two unseen crops, maize of Poaceae and potato of Solanaceae.
Results: LCFIN reaches 88.99% accuracy on maize and 95.23% on potato under matched acquisition conditions. Across five independent group-aware partitions the in-domain accuracy is 99.93% ± 0.05%, with 19 misclassified test images among 27,937. A duplicate and near-duplicate audit bounds the accuracy inflation attributable to data redundancy at a few hundredths of a percentage point. Ablation across the five partitions shows the two domains to be functionally complementary rather than redundant, but the interaction is sub-additive, with a residual of 0.04 percentage points below the additive prediction.
Discussion: These findings are consistent with the learned features capturing phytopathological regularities, although this observation remains descriptive rather than causal. They support the hypothesis that functional-domain separation induces implicit feature disentanglement and enables capture of phytopathological signatures shared across the tested families, without demonstrating that they transcend taxonomic boundaries. Because the extractor is frozen and only a linear head is fitted, extending the model to a further crop costs far less than training a network from scratch.
Related
- MSCF-DM: center-site-aware attention and gated multi-scale fusion network for cross-species DNA methylation site prediction
- FVRS: a cross-species multi-omics database for fruit and vegetable ripening and senescence
- Style transfer mapping with ensemble learning for cross-subject EEG-based neuropsychiatric disease recognition.
- ISP Platform powered by Geneformer: Framework for Cross-Species, Sequential, and Multi-Gene In Silico Perturbation Screens with Application to iPS Cell State Transitions
- Cross-Species and Cell-Type-Specific H3K27ac Dynamics Link Brain Aging to Alzheimer’s Disease Genes
BibTeX
@article{cai2026crosscrop,
title = {{Cross-crop generalization in multi-species leaf disease recognition}},
author = {Zhaopeng Cai and Nadia Farhana and Jiawei Li and Asif Mahbub Karim and Lizheng Guo and Junhui Cao},
journal = {Frontiers in Plant Science},
year = {2026},
month = oct,
doi = {10.3389/fpls.2026.1970636},
url = {https://www.frontiersin.org/articles/10.3389/fpls.2026.1970636/full},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Cross-crop generalization in multi-species leaf disease recognition
AU - Zhaopeng Cai
AU - Nadia Farhana
AU - Jiawei Li
AU - Asif Mahbub Karim
AU - Lizheng Guo
AU - Junhui Cao
PY - 2026
DA - 2026/10/29
JO - Frontiers in Plant Science
DO - 10.3389/fpls.2026.1970636
UR - https://www.frontiersin.org/articles/10.3389/fpls.2026.1970636/full
AB - Introduction: Plant diseases are major biotic stress factors whose visual symptoms may be conserved across taxonomic families, though the extent remains unquantified. Cross-crop leaf disease recognition requires architectures that generalize to species absent from training, but existing methods lack a theoretically grounded criterion for ensuring cross-species transferability. Methods: This study proposes functional-domain separation as such a criterion, derived from the disentangled representation learning hypothesis, and implements it as the leaf-disease composite feature interaction network (LCFIN), a configurable multi-branch feature enhancement framework. LCFIN assigns a two-level Squeeze-and-Excitation mechanism to a channel calibration domain and an MLP-Mixer head to a spatial interaction domain on an EfficientNet-B3 backbone, with dual-weighted sampling addressing cross-level class imbalance. It is trained for main fruit and vegetable leaf disease identification (MFVLDI) on three families (Rosaceae, Vitaceae, Solanaceae) and evaluated under a frozen-representation linear-probe protocol on two unseen crops, maize of Poaceae and potato of Solanaceae. Results: LCFIN reaches 88.99% accuracy on maize and 95.23% on potato under matched acquisition conditions. Across five independent group-aware partitions the in-domain accuracy is 99.93% ± 0.05%, with 19 misclassified test images among 27,937. A duplicate and near-duplicate audit bounds the accuracy inflation attributable to data redundancy at a few hundredths of a percentage point. Ablation across the five partitions shows the two domains to be functionally complementary rather than redundant, but the interaction is sub-additive, with a residual of 0.04 percentage points below the additive prediction. Discussion: These findings are consistent with the learned features capturing phytopathological regularities, although this observation remains descriptive rather than causal. They support the hypothesis that functional-domain separation induces implicit feature disentanglement and enables capture of phytopathological signatures shared across the tested families, without demonstrating that they transcend taxonomic boundaries. Because the extractor is frozen and only a linear head is fitted, extending the model to a further crop costs far less than training a network from scratch.
SN - 1664-462X
ER -