Scholarlink

The proposed work achieved a precision of 0.9538, recall of 0.9240, and an accuracy of 0.9908 on the Pascal-VOC-2012 dataset, demonstrating the superior ability of the proposed ensemble model to classify complex multi-label cases effectively. The model combines VGG19, ResNet101, and MobileNetV3 in an unweighted ensemble, leveraging the unique strengths of each architecture to enhance classification accuracy. Building upon the proven efficacy of deep convolutional networks in image classification tasks, such as ResNet, VGG, and MobileNet, this research extends these architectures to multi-label scenarios. It aligns with advancements in dataset design like Pascal-VOC-2012 and image processing innovations such as the Inception network, laying the groundwork for future improvements in multi-label ensemble learning for resource-constrained and real-time applications.

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

Multi-label Image Classification Using an Unweighted Ensemble

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Saarthak Sooji,
  • Nandan Kamaraddi,
  • Tarun Ejanthkar,
  • Chirag Shetty,
  • Uday Kulkarni,
  • Shashank Hegde

摘要

The proposed work achieved a precision of 0.9538, recall of 0.9240, and an accuracy of 0.9908 on the Pascal-VOC-2012 dataset, demonstrating the superior ability of the proposed ensemble model to classify complex multi-label cases effectively. The model combines VGG19, ResNet101, and MobileNetV3 in an unweighted ensemble, leveraging the unique strengths of each architecture to enhance classification accuracy. Building upon the proven efficacy of deep convolutional networks in image classification tasks, such as ResNet, VGG, and MobileNet, this research extends these architectures to multi-label scenarios. It aligns with advancements in dataset design like Pascal-VOC-2012 and image processing innovations such as the Inception network, laying the groundwork for future improvements in multi-label ensemble learning for resource-constrained and real-time applications.