Crop classification of large-scale agricultural land is crucial for crop monitoring and yield estimation. Hyperspectral image classification has proven to be an effective method for this task. Most current popular hyperspectral image classification methods are based on image classification, specifically on convolutional neural networks (CNNs) and recurrent neural networks (RNNs). In contrast, this paper focuses on methods based on semantic segmentation and proposes a new transformer-based approach called HyperSFormer for crop hyperspectral image classification. The key enhancement of the proposed method is the replacement of the encoder in SegFormer with an improved Swin Transformer while keeping the SegFormer decoder. The entire model adopts a simple and uniform transformer architecture. Additionally, the paper introduces the hyper patch embedding (HPE) module to extract spectral and local spatial information from the hyperspectral images, which enhances the effectiveness of the features used as input for the model. To ensure detailed model processing and achieve end-to-end hyperspectral image classification, the transpose padding upsample (TPU) module is proposed for the model’s output. In order to address the problem of insufficient and imbalanced samples in hyperspectral image classification, the paper designs an adaptive min log sampling (AMLS) strategy and a loss function that incorporates dice loss and focal loss to assist model training. Experimental results using three public hyperspectral image datasets demonstrate the strong performance of HyperSFormer, particularly in the presence of imbalanced sample data, complex negative samples, and mixed sample classes. HyperSFormer outperforms state-of-the-art methods, including fast patch-free global learning (FPGA), a spectral–spatial-dependent global learning framework (SSDGL), and SegFormer, by at least 2.7% in the mean intersection over union (mIoU). It also improves the overall accuracy and average accuracy values by at least 0.9% and 0.3%, respectively, and the kappa coefficient by at least 0.011. Furthermore, ablation experiments were conducted to determine the optimal hyperparameter and loss function settings for the proposed method, validating the rationality of these settings and the fusion loss function.

Due to the challenges of pest detection in complex environments, this research introduces a lightweight network for tobacco pest identification leveraging enhancements in YOLOv8 technology. Using YOLOv8 large (YOLOv8l) as the base, the neck layer of the original network is replaced with an asymptotic feature pyramid network (AFPN) network to reduce model parameters. A SimAM attention mechanism, which does not require additional parameters, is incorporated to improve the model’s ability to extract features. The backbone network’s C2f model is replaced with the VoV-GSCSP module to reduce the model’s computational requirements. Experiments show the improved YOLOv8 model achieves high overall performance. Compared to the original model, model parameters and GFLOPs are reduced by 52.66% and 19.9%, respectively, while mAP@0.5 is improved by 1%, recall by 2.7%, and precision by 2.4%. Further comparison with popular detection models YOLOv5 medium (YOLOv5m), YOLOv6 medium (YOLOv6m), and YOLOv8 medium (YOLOv8m) shows the improved model has the highest detection accuracy and lightest parameters for detecting four common tobacco pests, with optimal overall performance. The improved YOLOv8 detection model proposed facilitates precise, instantaneous pest detection and recognition for tobacco and other crops, securing high-accuracy, comprehensive pest identification.

Detecting litchis in a complex natural environment is important for yield estimation and provides reliable support to litchi-picking robots. This paper proposes an improved litchi detection model named YOLOv5-litchi for litchi detection in complex natural environments. First, we add a convolutional block attention module to each C3 module in the backbone of the network to enhance the ability of the network to extract important feature information. Second, we add a small-object detection layer to enable the model to locate smaller targets and enhance the detection performance of small targets. Third, the Mosaic-9 data augmentation in the network increases the diversity of datasets. Then, we accelerate the regression convergence process of the prediction box by replacing the target detection regression loss function with CIoU. Finally, we add weighted-boxes fusion to bring the prediction boxes closer to the target and reduce the missed detection. An experiment is carried out to verify the effectiveness of the improvement. The results of the study show that the mAP and recall of the YOLOv5-litchi model were improved by 12.9% and 15%, respectively, in comparison with those of the unimproved YOLOv5 network. The inference speed of the YOLOv5-litchi model to detect each picture is 25 ms, which is much better than that of Faster-RCNN and YOLOv4. Compared with the unimproved YOLOv5 network, the mAP of the YOLOv5-litchi model increased by 17.4% in the large visual scenes. The performance of the YOLOv5-litchi model for litchi detection is the best in five models. Therefore, YOLOv5-litchi achieved a good balance between speed, model size, and accuracy, which can meet the needs of litchi detection in agriculture and provides technical support for the yield estimation and litchi-picking robots.

The rapid and accurate identification of citrus leaf diseases is crucial for the sustainable development of the citrus industry. Because citrus leaf disease samples are small, unevenly distributed, and difficult to collect, we redesigned the generator structure of FastGAN and added small batch standard deviations to the discriminator to produce an enhanced model called FastGAN2, which was used for generating citrus disease and nutritional deficiency (zinc and magnesium deficiency) images. The performance of the existing model degrades significantly when the training and test data exhibit large differences in appearance or originate from different regions. To solve this problem, we propose an EfficientNet-B5 network incorporating adaptive angular margin (Arcface) loss with the adversarial weight perturbation mechanism, and we call it EfficientNet-B5-pro. The FastGAN2 network can be trained using only 50 images. The Fréchet Inception Distance (FID) and Kernel Inception Distance (KID) are improved by 31.8% and 59.86%, respectively, compared to the original FastGAN network; 8000 images were generated using the FastGAN2 network (2000 black star disease, 2000 canker disease, 2000 healthy, 2000 deficiency). Only images generated by the FastGAN2 network were used as the training set to train the ten classification networks. Real images, which were not used to train the FastGAN2 network, were used as the test set. The average accuracy rates of the ten classification networks exceeded 93%. The accuracy, precision, recall, and F1 scores achieved by EfficientNet-B5-pro were 97.04%, 97.32%, 96.96%, and 97.09%, respectively, and they were 2.26%, 1.19%, 1.98%, and 1.86% higher than those of EfficientNet-B5, respectively. The classification network model can be successfully trained using only the images generated by FastGAN2, and EfficientNet-B5-pro has good generalization and robustness. The method used in this study can be an effective tool for citrus disease and nutritional deficiency image classification using a small number of samples.

Predicting student performance is a crucial research area in educational data mining. Student grades are influenced by various factors such as past academic performance, family background, and personal achievements. However, it is difficult to collect many relevant factors other than past academic performance, and it is not ideal to predict student performance using past academic records alone. Therefore, reducing the difficulty of data collection while maintaining timeliness and prediction accuracy remains a pressing issue. To address this challenge, this study proposes a novel model for predicting university student performance based on multi-feature fusion and attention mechanisms. The proposed model focuses on analyzing historical academic grades from multiple dimensions among university students to extract features that reveal relationships between courses and students, among different students themselves, or even between courses themselves. Additionally, an attention mechanism is introduced to explore the relationship between different dimensional features. This study collected a triplet set of related courses and students’ real historical grades, proved the correlation between courses through data analysis, and verified the effectiveness of different dimensional features. Experimental results show that, compared with traditional machine learning methods, our proposed method achieves better prediction accuracy with a precision rate of 72.5%.

The determination of crop water status has positive effects on the Chinese Brassica industry and irrigation decisions. Drought can decrease the production of Chinese Brassica, whereas over-irrigation can waste water. It is desirable to schedule irrigation when the crop suffers from water stress. In this study, a random forest model was developed using sample data derived from meteorological measurements including air temperature (Ta), relative humidity (RH), wind speed (WS), and photosynthetic active radiation (Par) to predict the lower baseline ($T_{wet}$) and upper baseline ($T_{dry}$) canopy temperatures for Chinese Brassica from 27 November to 31 December 2020 (E1) and from 25 May to 20 June 2021 (E2). Crop water stress index (CWSI) values were determined based on the predicted canopy temperature and used to assess the crop water status. The study demonstrated the viability of using a random forest model to forecast $T_{wet}$ and $T_{dry}$. The coefficients of determination (R²) in E1 were 0.90 and 0.88 for development and 0.80 and 0.77 for validation, respectively. The R² values in E2 were 0.91 and 0.89 for development and 0.83 and 0.80 for validation, respectively. Our results reveal that the measured and predicted CWSI values had similar R² values related to stomatal conductance (~0.5 in E1, ~0.6 in E2), whereas the CWSI showed a poor correlation with transpiration rate (~0.25 in E1, ~0.2 in E2). Finally, the methodology used to calculate the daily CWSI for Chinese Brassica in this study showed that both $T_{wet}$ and $T_{dry}$, which require frequent measuring and design experiment due to the trial site and condition changes, have the potential to simulate environmental parameters and can therefore be applied to conveniently calculate the CWSI.