Scholarlink

<p>In recent years, the increasing prevalence of digital imagery has heightened the need for robust methods to authenticate and analyze images. The proposed framework focuses on identifying image duplications and applies advanced image processing techniques to boost the accuracy and efficiency of reconstructing phylogeny trees. Selected for its cultural diversity, the Indian Movie Face collection offers a unique set of challenges and opportunities for analysis. Image phylogeny, which involves tracing the evolutionary relationships among multiple images, is becoming increasingly vital in digital forensics and image analysis. Various transformation methods, including Cosine, Gaussian, Wavelet, and Fourier transforms, are utilized to enhance the framework. The performance of this framework is benchmarked against existing methods using standard metrics such as accuracy, precision, recall, and the F1 score. Results highlight the framework’s effectiveness, particularly in addressing the unique challenges presented by the Indian Movie Face and Labelled Faces in the Wild collections.</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

Biometric face phylogeny tree reconstruction using optimal texture extraction model

  • 期刊论文

  • 发表日期：

    2025年2月24日

查看全文
  • Hemalata Mote,
  • Sujata Kulkarni

摘要

In recent years, the increasing prevalence of digital imagery has heightened the need for robust methods to authenticate and analyze images. The proposed framework focuses on identifying image duplications and applies advanced image processing techniques to boost the accuracy and efficiency of reconstructing phylogeny trees. Selected for its cultural diversity, the Indian Movie Face collection offers a unique set of challenges and opportunities for analysis. Image phylogeny, which involves tracing the evolutionary relationships among multiple images, is becoming increasingly vital in digital forensics and image analysis. Various transformation methods, including Cosine, Gaussian, Wavelet, and Fourier transforms, are utilized to enhance the framework. The performance of this framework is benchmarked against existing methods using standard metrics such as accuracy, precision, recall, and the F1 score. Results highlight the framework’s effectiveness, particularly in addressing the unique challenges presented by the Indian Movie Face and Labelled Faces in the Wild collections.