Bruno Antonio, Moroni Davide, Dainelli Riccardo, Rocchi Leandro, Morelli Silvia, Ferrari Emilio, Toscano Piero, Martinelli Massimo
Institute of Information Science and Technologies, National Research Council, Pisa, Italy.
Institute of BioEconomy, National Research Council, Firenze, Italy.
Front Artif Intell. 2022 Sep 8;5:868926. doi: 10.3389/frai.2022.868926. eCollection 2022.
A novel method for improving plant disease classification, a challenging and time-consuming process, is proposed. First, using as baseline EfficientNet, a recent and advanced family of architectures having an excellent accuracy/complexity trade-off, we have introduced, devised, and applied refined techniques based on transfer learning, regularization, stratification, weighted metrics, and advanced optimizers in order to achieve improved performance. Then, we go further by introducing adaptive minimal ensembling, which is a unique input to the knowledge base of the proposed solution. This represents a leap forward since it allows improving the accuracy with limited complexity using only two EfficientNet-b0 weak models, performing ensembling on feature vectors by a trainable layer instead of classic aggregation on outputs. To the best of our knowledge, such an approach to ensembling has never been used before in literature. Our method was tested on PlantVillage, a public reference dataset used for benchmarking models' performances for crop disease diagnostic, considering both its original and augmented versions. We noticeably improved the state of the art by achieving 100% accuracy in both the original and augmented datasets. Results were obtained using PyTorch to train, test, and validate the models; reproducibility is granted by providing exhaustive details, including hyperparameters used in the experimentation. A Web interface is also made publicly available to test the proposed methods.
提出了一种改进植物病害分类的新方法，这是一个具有挑战性且耗时的过程。首先，以EfficientNet作为基线，EfficientNet是一个近期的先进架构家族，具有出色的准确率/复杂度权衡，我们引入、设计并应用了基于迁移学习、正则化、分层、加权度量和先进优化器的精细技术，以实现性能提升。然后，我们进一步引入自适应最小集成，这是所提出解决方案知识库的独特输入。这代表了一个飞跃，因为它仅使用两个EfficientNet - b0弱模型就能以有限的复杂度提高准确率，通过一个可训练层对特征向量进行集成，而不是对输出进行经典聚合。据我们所知，这种集成方法在文献中从未被使用过。我们的方法在PlantVillage上进行了测试，PlantVillage是一个用于对作物病害诊断模型性能进行基准测试的公共参考数据集，同时考虑了其原始版本和增强版本。我们通过在原始数据集和增强数据集中都达到100%的准确率，显著改进了现有技术水平。使用PyTorch训练、测试和验证模型获得了结果；通过提供详尽的细节（包括实验中使用的超参数）确保了可重复性。还公开提供了一个Web界面来测试所提出的方法。
