Publication

Abstract : In this paper, Using the ConvNeXt-Small architecture, we suggested a deep learning classification model in this experiment to categorize photos of papaya leaf diseases using the BDPapayaLeaf dataset. Images with annotations related to five distinct kinds of papaya plant diseases make up the BDPapayaLeaf dataset. The project’s main goal is to accurately diagnose these diseases using CNNs in order to protect plants, detect diseases early, and increase agricultural yield. according to the transfer learning feature. The ConvNeXt-Small model was optimized to meet our particular purpose after being trained in each dataset. To improve feature extraction from input photos, the final model was enhanced with additional convolutional and max-pooling layers. Techniques for data augmentation were applied to prevent overfitting and enhance the model’s capacity for generalization. The end model demonstrated its effectiveness in correctly classifying the five categories of papaya leaf diseases and demonstrated strong performance on both the training and validation datasets. In-depth discussions of dataset preprocessing, model architecture, training, assessment, and outcomes are provided in this work, which also identifies promising applications of deep learning techniques for agricultural disease categorization.