A Lightweight Convolutional Neural Network with Neighbourhood Attention and a 100- Category Dataset for Plant Disease Detection

Abstract

Plant disease detection is vital for agricultural sustainability and food security. While Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) have achieved high accuracy in this domain, CNNs often require millions of parameters and substantial computation. ViTs suffer from the quadratic time and space complexity of self-attention (SA), limiting their use on resource-constrained devices. Although SA is capable of modelling long-range dependencies when symptoms are dispersed, many plant diseases exhibit small, localized lesions or texture changes; therefore, Neighborhood Attention (NA) offers a more efficient and targeted alternative by focusing on nearby regions rather than the entire image. This work proposes a custom Localized NA block implemented in TensorFlow/Keras that operates directly on CNN feature maps, bypassing patch embedding and transformer modules. A lightweight CNN is then developed by combining depth-wise separable convolutions with the proposed localized NA block. In addition, a 100-category plant disease dataset covering 16 crops is presented. The dataset is curated, class-balanced, and made publicly available to support reproducibility and encourage further research. The proposed 9-layer CNN, with just 1.7M parameters and a size of 6.74 MB, achieved a favorable balance between accuracy, model size, and computational efficiency, compared with MobileNetV1, MobileNetV2, DenseNet121, InceptionV3, MobileViT-XXS, and EfficientViT-M0, achieving 98.97%± 0.33% accuracy on PlantVillage and 93.36%± 0.28% on the proposed dataset. The ablation study showed that the NA block improved test accuracy by approximately 2–3%, while Grad-CAM visualizations indicated more precise targeting of diseased areas in the leaf image.

Plant phenotyping relevance

植物葉画像から病徴を推定する軽量CNNと注意機構を開発し、複数データセットで比較評価・アブレーションを行い、さらに100カテゴリの公開データセットを提示しているため、植物フェノタイピング手法が中心である。

Code and data availability

The paper's authors curated a 100-category plant disease dataset and explicitly state it is publicly available on Kaggle in both augmented-train and raw split forms. These are paper-specific, public, actionable phenotype image datasets. The PlantVillage benchmark is a third-party dataset, not a paper-specific asset, so