Scholarlink

With the advancement of deep learning technology, facial forgery techniques have continuously improved, and detection methods have also evolved. However, most existing approaches utilize highly complex neural networks to mitigate overfitting on a single dataset, resulting in significant computational overhead. This paper proposes a feature extraction method based on image gradients, where gradient maps in various directions are computed to capture edge information and texture features. These gradient maps are then input into a neural network for classification, reducing computational costs while minimizing the risk of overfitting to a single dataset.

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 Detection of Face Forgery Using Intra-pixel Gradient Information

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Shiyu Chen,
  • Munkhdelgerekh Batzorig,
  • Eunseok Kim,
  • Purevbaatar Ganbold,
  • Kangbin Yim

摘要

With the advancement of deep learning technology, facial forgery techniques have continuously improved, and detection methods have also evolved. However, most existing approaches utilize highly complex neural networks to mitigate overfitting on a single dataset, resulting in significant computational overhead. This paper proposes a feature extraction method based on image gradients, where gradient maps in various directions are computed to capture edge information and texture features. These gradient maps are then input into a neural network for classification, reducing computational costs while minimizing the risk of overfitting to a single dataset.