Scholarlink

We proposed an attention-enhanced MobileNetV2 deep learning approach for the diagnosis of monkeypox accurately without overlapping symptoms of chickenpox and measles. A lightweight model was used, ideal for usage on mobile devices, operating in low-resource settings, and integrated with spatial and channel attention for higher accuracy detection. Added to compensate for the low images available of monkeypox were the chickenpox and measles images added to the data set. The model has been shown to be more efficient than ResNet, VGG, and AlexNet since it achieved an accuracy of 98.78% in the extended data set and 92% in the original MSID. Grad-CAM and LIME were the techniques used for interpretability so that healthcare professionals could trust the decisions made by the model. Thus, attention-augmented MobileNetV2 provides a high accuracy and interpretable solution for early disease detection in resource-limited environments.

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

Efficient Monkeypox Diagnosis Via Attention-Based MobileNetV2 Architecture

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Mothe Suneetha,
  • Shaik Khaja Muneer,
  • Surugula Srikanth,
  • Singu Vamsi,
  • R. Sathees Kumar,
  • Sireesha Moturi,
  • Dodda Venkata Reddy

摘要

We proposed an attention-enhanced MobileNetV2 deep learning approach for the diagnosis of monkeypox accurately without overlapping symptoms of chickenpox and measles. A lightweight model was used, ideal for usage on mobile devices, operating in low-resource settings, and integrated with spatial and channel attention for higher accuracy detection. Added to compensate for the low images available of monkeypox were the chickenpox and measles images added to the data set. The model has been shown to be more efficient than ResNet, VGG, and AlexNet since it achieved an accuracy of 98.78% in the extended data set and 92% in the original MSID. Grad-CAM and LIME were the techniques used for interpretability so that healthcare professionals could trust the decisions made by the model. Thus, attention-augmented MobileNetV2 provides a high accuracy and interpretable solution for early disease detection in resource-limited environments.