,CNN,使

Natthaphong Suthamno1, Jessada Tanthanuch1

  • 1School of Mathematical Sciences and Geoinformatics, Institute of Science, Suranaree University of Technology, Nakhon Ratchasima 30000, Thailand.

Journal of imaging
|February 26, 2026
PubMed
概括

这项研究通过使用主题建模来增强图像分类,在训练卷积神经网络 (CNN) 之前从语义上分组图像. 与标准方法相比,这种方法显著提高了分类准确性.

相关概念视频