AI SummaryAI-generated · verify against the full text
Deep learning models struggle to accurately classify maize roots due to image noise and class imbalance.
Researchers evaluated standard computer vision pipelines and deep learning models for classifying maize roots. They found that automated background removal methods often misinterpret fine root hairs as noise, leading to significant data loss. Additionally, widely used CNN architectures showed a strong bias towards the majority class, resulting in poor recall for the haploid class, even under realistic class imbalance conditions.
Abstract
Doubled Haploid (DH) technology significantly accelerates the development of homozygous lines in maize breeding; however, its scalability is constrained by the reliable discrimination of haploid and diploid individuals. The widely used R1-nj anthocyanin marker at the seed stage is susceptible to genetic suppression and environmental variability, leading to high misclassification rates. This limitation has driven a shift toward seedling root morphology as a more robust phenotypic marker, yet it introduces major challenges, including complex image noise and severe class imbalance. In this study, we systematically evaluate the limitations of standard computer vision pipelines and baseline deep learning models for root-based classification. Automated background removal methods (HSV, Rembg) are shown to misinterpret fine root hairs as noise, resulting in significant morphological data loss. Additionally, experiments conducted under a realistic class imbalance (1:5.4) demonstrate that widely used CNN architectures (ResNet50, VGG16, EfficientNetB0, DenseNet121) exhibit strong majority class bias, with haploid recall dropping to 0.00% and 27.7%. These findings reveal a critical limitation in existing approaches and highlight the need for domain-informed datasets and imbalance-aware learning strategies for robust and scalable AI-based maize breeding systems.
Keywords
Deep LearningClass ImbalanceImage PreprocessingMinority Class DetectionConvolutional Neural Networks (CNN)Root PhenotypingAgricultural Image Analysis
Full TextOpen Access
References
- 1.[1] Y. Altuntaş, Z. Cömert, and A. F. Kocamaz, "Identification of haploid and diploid maize seeds using convolutional neural networks and a transfer learning approach," Comput. Electron. Agric., vol. 163, p. 104874, 2019. doi: 10.1016/j.compag.2019.104874.
- 2.[2] B. W. Boote, D. J. Freppon, G. N. De La Fuente, T. Lübberstedt, B. J. Nikolau, and E. A. Smith, "Haploid differentiation in maize kernels based on fluorescence imaging," Plant Breed., vol. 135, no. 4, pp. 439-445, 2016. doi: 10.1111/pbr.12382.
- 3.[3] M. Buda, A. Maki, and M. A. Mazurowski, "A systematic study of the class imbalance problem in convolutional neural networks," Neural Networks, vol. 106, pp. 249-259, 2018. doi: 10.1016/j.neunet.2018.07.011.
- 4.[4] V. Chaikam, W. Molenaar, A. E. Melchinger, and P. M. Boddupalli, "Doubled haploid technology for line development in maize: technical advances and prospects," Theor. Appl. Genet., vol. 132, no. 12, pp. 3227-3243, 2019. doi: 10.1007/s00122-019-03433-x.
- 5.[5] D. Chicco and G. Jurman, "The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation," BMC Genomics, vol. 21, no. 1, pp. 1-13, 2020. doi: 10.1186/s12864-019-6413-7.
- 6.[6] A. Dermail, M. Mitchell, T. Foster, M. Fakude, Y. R. Chen, K. Suriharn, U. K. Frei, and T. Lübberstedt, "Haploid identification in maize," Front. Plant Sci., vol. 15, p. 1378421, 2024. doi: 10.3389/fpls.2024.1378421.
- 7.[7] H. H. Geiger and G. A. Gordillo, "Doubled haploids in hybrid maize breeding," Maydica, vol. 54, no. 4, p. 485, 2009.
- 8.[8] He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 770-778).
- 9.[9] Y. Ho and S. Wookey, "The real-world-weight cross-entropy loss function: Modeling the costs of misclassification," IEEE Access, vol. 8, pp. 4806-4813, 2019. doi: 10.1109/ACCESS.2019.2962617.
- 10.[10] Huang, G., Liu, Z., Van Der Maaten, L., & Weinberger, K. Q. (2017). Densely connected convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 4700-4708).
- 11.[11] F. Kahrıman, A. M. Güz, and İ. Pehlivan, "Use of machine learning models-based image analysis to discriminate haploid and diploid seeds in maize," Sci. Agric., vol. 80, 2023. doi: 10.1590/1984-70332023v23n4a44.
- 12.[12] W. Liao, X. Wang, D. An, and Y. Wei, "Hyperspectral imaging technology and transfer learning utilized in identification of haploid maize seeds," Sensors, vol. 19, no. 5, p. 1167, 2019. doi: 10.48550/arXiv.1805.11784.
- 13.[13] J. Pace, N. Lee, H. S. Naik, B. Ganapathysubramanian, and T. Lübberstedt, "Analysis of maize (Zea mays L.) seedling roots with the high-throughput image analysis tool ARIA (Automatic Root Image Analysis)," PLoS ONE, vol. 9, no. 9, p. e108255, 2014. doi: 10.1371/journal.pone.0108255.
- 14.[14] M. P. Pound, J. A. Atkinson, A. J. Townsend, M. H. Wilson, M. Griffiths, A. S. Jackson, et al., "Deep machine learning provides state-of-the-art performance in image-based plant phenotyping," Gigascience, vol. 6, no. 10, p. gix083, 2017. doi: 10.1093/gigascience/gix083.
- 15.[15] B. M. Prasanna, V. Chaikam, and G. Mahuku, *Doubled Haploid Technology in Maize Breeding: Theory and Practice*. CIMMYT, 2012.
- 16.[16] T. Saito and M. Rehmsmeier, "The precision-recall plot is more informative than the ROC plot when evaluating binary classifiers on imbalanced datasets," PLoS ONE, vol. 10, no. 3, p. e0118432, 2015. doi: 10.1371/journal.pone.0118432.
- 17.[17] K. Simonyan and A. Zisserman, "Very deep convolutional networks for large-scale image recognition," arXiv preprint arXiv:1409.1556, 2014.
- 18.[18] Tan, M., & Le, Q. (2019). EfficientNet: Rethinking model scaling for convolutional neural networks. In International conference on machine learning (pp. 6105-6114). PMLR.
- 19.[19] B. Veeramani, J. W. Raymond, and P. Chanda, "DeepSort: deep convolutional networks for sorting haploid maize seeds," BMC Bioinformatics, vol. 19, no. 9, pp. 1-10, 2018. doi: 10.1186/s12859-018-2267-2.
- 20.[20] X. Y. Wang, W. X. Liao, D. An, and Y. G. Wei, "Maize haploid identification via LSTM-CNN and hyperspectral imaging technology," arXiv preprint arXiv:1805.09105, 2018.
- 21.[21] L. Zhang, D. Wang, J. Liu, and D. An, "Vis-NIR hyperspectral imaging combined with incremental learning for open world maize seed varieties identification," Front. Plant Sci., vol. 13, p. 893630, 2022. doi: 10.3389/fpls.2022.893630.