Scholarlink

<p>Fruit disease detection prevents fruit production losses and contributes to strengthening the national economy. In the past few decades, manual detection has played a&#xa0;major role in fruit disease detection; nevertheless, monitoring fruits manually is an extremely difficult and time-consuming process. Thus, machine learning (ML)-based models have emerged recently to solve time constraints. However, the ML models produce inaccurate results due to the lack of interpretation, generalization ability, and reliability. To address these challenges, this study proposes the Osphresis-flash searching optimization-based distributed triplet attention-enabled artificial neural network and light gradient boosting machine (OFlS-DTA2LM) model for fruit disease detection. The OFlS-DTA2LM model has powerful capabilities to handle complex and high-dimensional features with the distributed triplet attention mechanism, which enhances interpretability and provides maximum accuracy for effective detection. In addition, the OFlS optimization algorithm plays a&#xa0;crucial role in feature selection and tuning the hyperparameters of the proposed model, which improves convergence and minimizes overfitting issues. Moreover, the OFlS-DTA2LM model offers stability, reliability, and robustness of fruit disease detection and achieves 97.11% accuracy, 99.19% precision, 94.74% recall, and a&#xa0;96.91% F1&#xa0;score using the Mango Fruit DDS dataset when compared to other state-of-the-art methods.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

OFlS-DTA2LM: Distributed Triplet Attention-Enabled Machine Learning Model for Fruit Disease Detection from Fruit Images

  • 期刊论文

  • 发表日期：

    2025年11月3日

查看全文
  • Rakesh Suryawanshi,
  • Kailas Patil

摘要

Fruit disease detection prevents fruit production losses and contributes to strengthening the national economy. In the past few decades, manual detection has played a major role in fruit disease detection; nevertheless, monitoring fruits manually is an extremely difficult and time-consuming process. Thus, machine learning (ML)-based models have emerged recently to solve time constraints. However, the ML models produce inaccurate results due to the lack of interpretation, generalization ability, and reliability. To address these challenges, this study proposes the Osphresis-flash searching optimization-based distributed triplet attention-enabled artificial neural network and light gradient boosting machine (OFlS-DTA2LM) model for fruit disease detection. The OFlS-DTA2LM model has powerful capabilities to handle complex and high-dimensional features with the distributed triplet attention mechanism, which enhances interpretability and provides maximum accuracy for effective detection. In addition, the OFlS optimization algorithm plays a crucial role in feature selection and tuning the hyperparameters of the proposed model, which improves convergence and minimizes overfitting issues. Moreover, the OFlS-DTA2LM model offers stability, reliability, and robustness of fruit disease detection and achieves 97.11% accuracy, 99.19% precision, 94.74% recall, and a 96.91% F1 score using the Mango Fruit DDS dataset when compared to other state-of-the-art methods.