This dataset encompasses an extensive collection of 18,248 high-resolution JPEG images, documenting various stages of Taro Leaf Blight (TLB) infection in Taro plants across West Africa. TLB, primarily caused by the pathogen Phytophthora colocasiae, manifests through necrotic leaf spots, white sporangia bands, and orange droplets, severely impacting the agricultural output and economic stability of smallholder farmers in the region. The images represent a range of infection stages-early, mid, late, and healthy conditions-captured during the dry and early rainy seasons in Nigeria and Ghana using smartphones equipped with high-resolution cameras. This dataset was carefully curated to help in the development and training of machine learning models for early and accurate detection of TLB, a crucial step towards effective disease management. By enabling the application of advanced diagnostics through technologies such as smartphone apps and AI-based analysis tools, this dataset not only aims to enhance the technological capabilities within agricultural sectors but also serves as a vital educational resource. Researchers and developers can utilize this dataset to create and refine models that diagnose plant diseases promptly, thereby allowing for timely interventions that can prevent widespread crop damage and subsequent economic losses. Additionally, the dataset supports ongoing efforts to integrate artificial intelligence with traditional farming practices, offering a bridge between advanced technological solutions and accessible applications for resource-limited settings. The potential reuse of this dataset extends beyond disease identification; it encompasses agricultural research, educational purposes, and further development of automated systems for plant health monitoring, making it a cornerstone for future innovations in agricultural technology and management strategies.
Why it matches plant phenotyping methodsタロイモ葉の病害症状を画像で記録した大規模データセットであり、植物の病害状態を推定する画像ベース表現型解析の基盤として、データセット自体が中心的成果である。
abstractThis dataset encompasses an extensive collection of 18,248 high-resolution JPEG images, documenting various stages of Taro Leaf Blight (TLB) infection in Taro plants across West Africa.
Reproduction assets foundThe paper is a Data in Brief describing a public plant-phenotyping image dataset (18,248 taro leaf blight images) deposited on Mendeley Data with an explicit direct URL and DOI, matching an allowed URL exactly.Dataset · publiction
• Institution :
University of Lagos, Akoka.
Kwame Nkrumah University of Science and Technology
• City/Town/Region:
Abakaliki, Ebonyi, Izzi, Ezza North, Agbani, Ngwo, Ashanti.
• Country : Nigeria and Ghana
Data accessibility
Repository name: Mendeley Data
Data identification number: 10.17632/3knm93dkc5.1
Direct URL to data: https://data.mendeley.com/datasets/3knm93dkc5/1
Related research article
Nwaneto, C., Yiinka-Banjo, C., Ugot, O. A., Annor, T., & Umeugochukwu, O. (2024). EARLY DETECTION OF THE TARO LEAF BLIGHT DISEASE IN THE WEST AFRICAN SUB-REGION USING DEEP IMAGE CLASSIFICATION MODELS. Smart Agricultural Technology , 100,636.
1
Value of the Data
•
This dataset is important for devOpen asset ↗Mendeley Data · 10.17632/3knm93dkc5.1lines:1-60Code / dataset availability confirmedCrossref · checked 6 Sept 2026
Early diagnosis and preventive measures are necessary to mitigate diseases’ impact on the yield of Colocasia esculenta (Taro). This study addresses the challenges of Taro disease identification by employing two key strategies: integrating explainable artificial intelligence techniques to interpret deep learning models and conducting a comparative analysis of advanced architectures Inception V3, ResNet-50, and Vision Transformers for classifying common Taro diseases, including leaf blight and mosaic virus, as well as identifying healthy leaves. The novelty of this work lies in the first-ever integration of SHapley Additive exPlanations (SHAP) with deep learning architectures to enhance model interpretability while providing a comprehensive comparison of state-of-the-art methods for this underexplored crop. The proposed models significantly improve the ability to recognize complex patterns and features, achieving high accuracy and robust performance in disease classification. The model’s efficacy was evaluated through multi-class statistical metrics, including accuracy, precision, F1 score, recall, specificity, Chohen’s kappa, and area under the curve. Among the architectures, Inception V3 exhibited superior performance in accuracy (0.9985), F1 score (0.9985), recall (0.9985), and specificity (0.9992). The explainability of Inception V3 was further enhanced using SHAP, which provides insights by dissecting the contributions of individual features in Taro leaves to the model’s predictions. This approach facilitates a deeper understanding of the disease classification process and supports the development of effective disease management strategies, ultimately contributing to improved Taro cultivation practices.
Why it matches plant phenotyping methodsタロイモ葉の画像から病害状態を分類する深層学習手法を比較・評価し、SHAPによる解釈性も検証しており、植物表現型（病害状態）の取得・推定が中心である。
abstractconducting a comparative analysis of advanced architectures Inception V3, ResNet-50, and Vision Transformers for classifying common Taro diseases, including leaf blight and mosaic virus, as well as identifying healthy leaves
Reproduction assets foundThe paper's phenotyping input is the public Colocasia esculenta Leaf Image Dataset (2062 taro leaf images: healthy, leaf blight, mosaic virus) hosted on Mendeley Data, explicitly linked in the Data Availability Statement. No author analysis code or trained model checkpoints are disclosed; the figshare SHAP supplement'sDataset · publicData Availability Statement: The data presented in this study are available in Mendeley Data
at https://data.mendeley.com/datasets/hmdr3dz3v6/2, accessed on 24 December 2024, doi:
10.17632/hmdr3dz3v6.2.Open asset ↗Mendeley Data · 10.17632/hmdr3dz3v6.2pdf-page:16 lines:1-58