Scholarlink

<p>One of the most fatal diseases that affect people is skin cancer. Because nevus and melanoma lesions are so similar and there is a high likelihood of false negative diagnoses challenges in hospitals. The aim of this paper is to propose and develop a technique to classify type of skin cancer with high accuracy using minimal resources and lightweight federated transfer learning models. Here minimal resource based pre-trained deep learning models including EfficientNetV2S, EfficientNetB3, ResNet50, and NasNetMobile have been used to apply transfer learning on data of shape<InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2024_82402_Article_IEq1.gif" Format="GIF" Height="14" Rendition="HTML" Resolution="72" Type="Linedraw" Width="107" /> </InlineMediaObject> <EquationSource Format="TEX">\(\:\:224\times\:224\times\:3\)</EquationSource> </InlineEquation>. To compare with applied minimal resource transfer learning, same methodology has been applied using best identified model i.e. EfficientNetV2S for images of shape<InlineEquation ID="IEq2"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2024_82402_Article_IEq2.gif" Format="GIF" Height="14" Rendition="HTML" Resolution="72" Type="Linedraw" Width="92" /> </InlineMediaObject> <EquationSource Format="TEX">\(\:\:32\times\:32\times\:3\)</EquationSource> </InlineEquation>. The identified minimal and lightweight resource based EfficientNetV2S with images of shape <InlineEquation ID="IEq3"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2024_82402_Article_IEq3.gif" Format="GIF" Height="14" Rendition="HTML" Resolution="72" Type="Linedraw" Width="91" /> </InlineMediaObject> <EquationSource Format="TEX">\(\:32\times\:32\times\:3\)</EquationSource> </InlineEquation> have been applied for federated learning ecosystem. Both, identically and non-identically distributed datasets of shape <InlineEquation ID="IEq4"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2024_82402_Article_IEq4.gif" Format="GIF" Height="14" Rendition="HTML" Resolution="72" Type="Linedraw" Width="91" /> </InlineMediaObject> <EquationSource Format="TEX">\(\:32\times\:32\times\:3\)</EquationSource> </InlineEquation> have been applied and analyzed through federated learning implementations. The results have been analyzed to show the impact of low-pixel images with non-identical distributions over clients using parameters such as accuracy, precision, recall and categorical losses. The classification of skin cancer shows an accuracy of IID 89.83% and Non-IID 90.64%.</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

Minimal sourced and lightweight federated transfer learning models for skin cancer detection

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年1月21日

查看全文
  • Vikas Khullar,
  • Prabhjot Kaur,
  • Shubham Gargrish,
  • Anand Muni Mishra,
  • Prabhishek Singh,
  • Manoj Diwakar,
  • Anchit Bijalwan,
  • Indrajeet Gupta

摘要

One of the most fatal diseases that affect people is skin cancer. Because nevus and melanoma lesions are so similar and there is a high likelihood of false negative diagnoses challenges in hospitals. The aim of this paper is to propose and develop a technique to classify type of skin cancer with high accuracy using minimal resources and lightweight federated transfer learning models. Here minimal resource based pre-trained deep learning models including EfficientNetV2S, EfficientNetB3, ResNet50, and NasNetMobile have been used to apply transfer learning on data of shape \(\:\:224\times\:224\times\:3\) . To compare with applied minimal resource transfer learning, same methodology has been applied using best identified model i.e. EfficientNetV2S for images of shape \(\:\:32\times\:32\times\:3\) . The identified minimal and lightweight resource based EfficientNetV2S with images of shape \(\:32\times\:32\times\:3\) have been applied for federated learning ecosystem. Both, identically and non-identically distributed datasets of shape \(\:32\times\:32\times\:3\) have been applied and analyzed through federated learning implementations. The results have been analyzed to show the impact of low-pixel images with non-identical distributions over clients using parameters such as accuracy, precision, recall and categorical losses. The classification of skin cancer shows an accuracy of IID 89.83% and Non-IID 90.64%.