Scholarlink

<p>The proliferation of counterfeit products with fake logos poses significant risks to consumer trust and e-commerce platform integrity. Logos, often containing text, symbols, and numeric elements, is a critical identifier for authentic products. To address this, the Termite-based Convolutional Predicting System (TbCPS) is developed, incorporating feature analysis, forecasting parameters, and preprocessing layers to filter noise from input images. The refined data is used for feature extraction and accurate detection of fake or unauthorized logos<b>.</b> Evaluated on datasets containing genuine and counterfeit logos, TbCPS achieves 99.64% accuracy with a detection time of 1.50&#xa0;s per image<b>,</b> outperforming conventional detection models in accuracy, sensitivity, precision, error rate, and F1-score<b>.</b> The results demonstrate TbCPS as a robust, efficient, and scalable solution for automated logo verification in e-commerce environments.</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

An optimized convolutional network-based prediction apporach for fake logo detection in e-commerce products

  • 期刊论文

  • 发表日期：

    2026年2月25日

查看全文
  • Preeti C. M,
  • T. Santhi Sri

摘要

The proliferation of counterfeit products with fake logos poses significant risks to consumer trust and e-commerce platform integrity. Logos, often containing text, symbols, and numeric elements, is a critical identifier for authentic products. To address this, the Termite-based Convolutional Predicting System (TbCPS) is developed, incorporating feature analysis, forecasting parameters, and preprocessing layers to filter noise from input images. The refined data is used for feature extraction and accurate detection of fake or unauthorized logos. Evaluated on datasets containing genuine and counterfeit logos, TbCPS achieves 99.64% accuracy with a detection time of 1.50 s per image, outperforming conventional detection models in accuracy, sensitivity, precision, error rate, and F1-score. The results demonstrate TbCPS as a robust, efficient, and scalable solution for automated logo verification in e-commerce environments.