High-Accuracy Detection of Maize Leaf Diseases CNN Based on Multi-Pathway Activation Function Module

Highlights Improving the robustness of maize leaf disease classification models in complex environments using multiple data enhancement methods. Fusing ResNet50 with the triplet attention mechanism to improve feature extraction. Reducing the use of labeled data to some extent by using SimCLR. The classification accuracy of four different leaf types was 91.87%, which was higher than other compared models. Abstract. Deep learning methods for classifying maize leaf diseases often need a lot of labeled data for training during the model training phase, but labeling data is sometimes difficult and expensive. In addition, complicated environmental circumstances can readily impede the recognition effect. In order to solve this issue, SimCLR and the triplet attention mechanism were coupled to create a system for classifying maize leaf diseases. Firstly, to increase the robustness of the model, several data enhancement techniques were used to imitate the interference of complex components in the natural environment. Secondly, the use of labeled data was minimized by learning the similarity between related categories using a self-supervised SimCLR framework. The feature extraction network simultaneously employed a triplet attention mechanism to enhance the model’s ability to extract information about the interactions between space and channel, improve attention to discriminative features, and lessen interference from complex factors. In order to verify the effects of incorporating SimCLR and triplet attention mechanisms, this work used three different types of maize leaf diseases and healthy leaves as experimental objects. The final enhanced model was compared and analyzed with seven widely used image classification models. The experimental results demonstrated that the maize leaf disease classification model suggested in this article effectively enhanced the classification performance of maize leaf diseases containing complex factors of interference with the test set, with the network model proposed in this article having an average classification accuracy of 91.87%, better than the seven comparative models and requiring no additional labeled data during the training process. Keywords: Image classification, Maize leaf disease, Self-supervised learning, SimCLR, Triplet attention module.

Maize, a vital global food crop, is susceptible to various diseases impacting agricultural yield and quality, necessitating effective monitoring strategies. This study presents a comprehensive approach utilizing advanced Convolutional Neural Network (CNN) architecture for efficient maize leaf disease detection and classification. Leveraging datasets from Kaggle, Harvard Dataverse, and Mendeley, the proposed CNN maize disease detection (MDD) system integrates optimization techniques, data augmentation, batch size adjustments, and fine-tuning strategies. Evaluation through data splitting reveals significant correlations among maize leaf classifications and the influence of data augmentation on image quality. Following training, the CNN-MDD model is deployed on a Microcontroller Unit (MCU) simulation, achieving promising optimization parameters: latency of 7.60 ms, RAM size of 726.60 KB, classifier flash size of 344.70 KB, and an accuracy of 94.60 %. These optimized parameters align well with MCU specifications, highlighting the potential for practical deployment on low-cost edge devices like Arduino Ble 33 Sense. Additionally, this research contributes to the creation of a Ghanaian maize crop pest and disease dataset and the development of an optimal and cost-effective CNN-MDD MCU edge device, aiming to provide local farmers in Ghana and Sub-Saharan Africa with a swift, offline solution for maize disease detection, thus fostering sustainable agricultural practices. We conduct a comparative analysis of Edge Impulse and TensorFlow, two prominent machine-learning platforms for TinyML model development and deployment in various applications, including smart crop disease detection. Utilizing a dataset comprising 12,344 raw images and 7,454 augmented images of healthy and diseased maize leaves, we evaluate the workflow, features, performance, and usability of each platform. Our findings indicate that Edge Impulse holds a slight advantage over TensorFlow in terms of ease of use, reliability, and memory footprint, whereas TensorFlow offers greater flexibility, customization options, and higher accuracy. We conclude that while both platforms are suitable for crop disease detection, a combined approach allows for leveraging TensorFlow's robust training capabilities for complex neural networks alongside the simplicity of Edge Impulse for deploying models on low-powered edge devices. This synergistic approach enhances the effectiveness and efficiency of crop disease detection systems, ultimately benefiting agricultural practices and food security.

Maize leaf diseases significantly reduce maize yield; therefore, monitoring and identifying the diseases during the growing season are crucial. Some of the current studies are based on images with simple backgrounds, and the realistic field settings are full of background noise, making this task challenging. We collected low-cost red, green, and blue (RGB) images from our experimental fields and public dataset, and they contain a total of four categories, namely, southern corn leaf blight (SCLB), gray leaf spot (GLS), southern corn rust (SR), and healthy (H). This article proposes a model different from convolutional neural networks (CNNs) based on transformer and self-attention. It represents visual information of local regions of images by tokens, calculates the correlation (called attention) of information between local regions with an attention mechanism, and finally integrates global information to make the classification. The results show that our model achieves the best performance compared to five mainstream CNNs at a meager computational cost, and the attention mechanism plays an extremely important role. The disease lesions information was effectively emphasized, and the background noise was suppressed. The proposed model is more suitable for fine-grained maize leaf disease identification in a complex background, and we demonstrated this idea from three perspectives, namely, theoretical, experimental, and visualization.

Maize leaf disease recognition based on improved MSRCR and OSCRNet

Maize leaf disease image enhancement algorithm using TFEGAN

Plant disease management is an essential process to minimize loss in the field of agriculture. Plant leaf disease detection(PLDD) technology helps the formers to reduce the loss of quality yield production. This study focuses on detecting paddy and maize plant leaf disease-detecting methods. It contributes to a PLDD method to improve detection accuracy and overall performance. The PLDD method introduced a whales-optimized artificial neural network (WOANN) method for classifying the four Maize and four rice leaf disease-related classes. The WOANN uses the whale optimizer's food-searching functionalities to improve the performance and detection accuracy of the dense net model. This WOANN classifies the maize leaf light, maize grey leaf spot, maize common rust, rice bacterial leaf blight, rice bacterial leaf steak, rice brown spot, and healthy leaves of both Maize and paddy. It uses the Hilbert-Schmidt independent criterion lasso correlation algorithm to support the WOANN classifier in selecting the significant features. The performance analysis shows that the WOANN-based approach achieves a maximum of 99.35% detection accuracy for maize leaf disease and 99.13% for paddy leaf disease. Its efficiency analysis shows that the WOANN-based approaches achieve a maximum accuracy rate than comparison approaches.

Detecting pests and diseases on maize leaves is challenging. This is especially true under complex conditions, such as variable lighting and occlusion. Current methods suffer from low detection accuracy. They also lack sufficient real-time performance. Hence, this study introduces the lightweight detection method YOLOv11-RCDWD based on an improved YOLOv11 model. The proposed approach enhances the YOLOv11 model by incorporating the RepLKNet module as the backbone, which significantly enhances the model’s capacity to capture characteristics of maize leaf pests and diseases. Additionally, the CBAM is embedded within the neck feature extraction network to further refine the feature representation to augment the model’s capability to identify and select essential features by introducing attention mechanisms in both the channel and spatial dimensions, thereby improving the accuracy of feature expression. We have also improved the model by incorporating the DynamicHead module, WIoU loss function, and DynamicATSS label assignment strategy, which collectively enhance detection accuracy, efficiency, and robustness through optimized attention mechanisms, better handling of low-quality samples, and dynamic sample selection during training. The experimental findings indicate that the improved YOLOv11-RCDWD model effectively detected pests and diseases on maize leaves. The precision reached 92.6%, while the recall was 85.4%. The F1 score was 88.9%, and the mAP@0.5 and mAP@0.5~0.95 demonstrated an improvement of 4.9% and 9.0% over the baseline YOLOv11s. Notably, the YOLOv11-RCDWD model significantly outperformed other architectures such as Faster R-CNN, SSD, and various models within the YOLO series, demonstrating superior capabilities in terms of detection speed, parameter count, computational efficiency, and memory utilization. This model achieves an optimal balance between detection performance and resource efficiency. Overall, the improved YOLOv11-RCDWD model significantly reduces detection time and memory usage while maintaining high detection accuracy, supporting the automated detection of maize pests and diseases, and offering a robust solution for intelligent monitoring of agricultural pests.

Getting accurate lifetime predictions for a particular cell chemistry remains a challenging process, largely dependent on time and cost-intensive experimental battery testing. This paper proposes a transfer learning (TL) method to develop LIB ageing models, which allow for the leveraging of experimental laboratory testing data previously obtained for a different cell technology. The TL method is implemented through Neural Networks models, using LiNiMnCoO2/C laboratory ageing data as a baseline model. The obtained TL model achieves an 1.01% overall error for a broad range of operating conditions, using for retraining only two experimental ageing tests of LiFePO4/C cells.

Transfer learning using computational intelligence: A survey

This work proposes a deep learning approach for maize leaf disease detection and classification with optimized techniques in Convolutional Neural Networks. The adopted dataset includes images of diverse maize leaf diseases categorized into four classes: Blight, Common Rust, Gray Leaf Spot, and Healthy, comprising 1,200 training samples, 400 validation samples, and 398 testing samples. The architecture of the model is designed to have convolutional layers, pooling layers, and dense layers for extracting features, while final classification is done with a softmax activation function. Advanced optimization techniques such as Adam optimizer, early stopping, and data augmentation were used to have the best generalization performance. Evaluation metrics include accuracy, loss, precision, recall, F1-score, and ROC-AUC, achieving an overall test accuracy of 93% with an AUC score of 0.99. The results indicate a robust solution for the accurate and scalable diagnosis of maize disease, with possible applications in agricultural monitoring systems.

Dynamic multiobjective optimization problems (DMOPs) aim to optimize multiple (often conflicting) objectives that are changing over time. Recently, there are a number of promising algorithms proposed based on transfer learning methods to solve DMOPs. However, it is very challenging to reduce the negative effect in transfer learning and find more effective transferred solutions. To fill this research gap, this article proposes a clustering-based transfer (CBT) learning method to solve DMOPs. When the environment changes, two novel operations (clustering-based selection (CBS) and CBT) are used to guide knowledge transfer. Specifically, CBS aims to find a population with nondominated solutions and dominated solutions as the training data for the new environment. Then, CBT further collects the previous Pareto-optimal solutions and some noise solutions as the training data for the previous environment. Two training data sets from different environments are, respectively, divided into multiple clusters and transfer learning is conducted on two similar clusters with high probability to reduce the negative effect, which can train an accurate prediction model to identify the promising solutions for the new environment. Empirical studies have been conducted on 14 benchmark DMOPs and one real-life path planning problem of unmanned air/ground vehicles, which validate the effectiveness of our proposed method. Especially, our method can significantly reduce negative transfer on 12 out of 14 cases when compared with direct transfer learning.

Surface electromyogram (sEMG)-based human-computer interface (HCI) is an effective tool for detecting human movements. Because sEMG-based motion recognition usually requires prolonged data measurements from the user (target), transfer learning reusing pre-measured (source) data from other users and pre-trained classifiers can be applied to sEMG data to reduce the measurement time. However, little knowledge is available regarding the combination of transfer learning methods and classifiers in sEMG data applications. Thus, we investigated the classification accuracy of data- and parameter-space-based transfer learning with shallow or deep classifiers in cross-subject sEMG classification. The dataset contains eight classes of forearm motions recorded from 25 volunteer participants. We used a support vector machine (SVM) as a shallow classifier as well as a deep neural network architecture, referred to as an artificial neural network (ANN), as a deep classifier. In addition, we used style transfer mapping (STM) as a data-space-based transfer learning method and fine-tuning (FT) as a parameter-space-based transfer learning method. Consequently, the classification accuracy of the ANN was higher than that of the SVM, regardless of the combinational use of transfer learning. STM and FT significantly improved the classification accuracy compared with non-transfer cases regardless of the classifier (note that FT can only be used with the ANN). In particular, the combined use of FT and the ANN yielded the best accuracy. These findings suggest that parameter-space-based transfer learning and deep classifiers are suitable for cross-subject sEMG classification. The combined use of parameter-space-based transfer learning and deep classifiers can effectively reduce the data measurement time of sEMG-based HCI applications.

Regression analysis of interval-censored failure time data commonly arises in biomedical studies, particularly when the available sample size is limited. Although many methods have been proposed for the semiparametric probit model with interval-censored data, there does not appear to exist an established approach that effectively borrows information from external sources to improve estimation efficiency. Such external information may arise, for example, in clinical trials where an auxiliary dataset from a related population is available but may differ from the target population in certain aspects, leading to heterogeneity between populations. To address this issue, a sieve maximum likelihood estimation procedure is developed for the semiparametric probit model with interval-censored data, and a transfer learning method is proposed to leverage auxiliary information from a source domain to improve estimation efficiency in the target domain while accounting for population heterogeneity. The proposed approach is based on a penalized likelihood formulation and uses monotone splines to approximate the unknown baseline function, providing flexibility in both modeling and computation. Simulation studies show that the proposed estimator substantially improves estimation accuracy compared with methods that rely solely on the target data, particularly when the target sample size is small. An application to an Alzheimer’s disease dataset further illustrates the practical usefulness of the proposed approach in biomedical studies.

Thermal imaging has been used for decades to monitor the health status of neonates as an non-invasive and non-ionizing imaging technique. Applications such as thermal asymmetry and disease analysis can be performed by applying deep learning methods to thermal imaging technique. However, thousands of different images are needed to perform analyzes with deep learning methods. It takes many years to create data sets with thousands of different images due to feeding time, medication time and instant baby care in the neonatal intensive care unit. In this study, a unhealthy-healthy classification was performed using thermal images obtained from the Selcuk University, Faculty of Medicine, Neonatal Intensive Care Unit for one year. Transfer learning method has been used to overcome the lack of data problem. When VGG16 model was used for transfer learning, the results were obtained as 100% sensitivity and 94.73% specificity. This result shows that thermal imaging and transfer learning method can be used in early diagnosis of diseases.

Neural transfer learning for assigning diagnosis codes to EMRs