Scholarlink

This work presents a new methodology for generating synthetic images. This methodology is based on a computational model of the Visual Cortex and employs multi-tree genetic programming to optimize the process. Three metrics are proposed as a fitness function to synthesize an image and are compared with each other. Likewise, the number of operations used in the process is estimated and compared with those used by a typical DCGAN, concluding that our proposal requires fewer operations.

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

Digital Image Synthesis Using Multi-tree Genetic Programming

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Eddie Clemente,
  • Ismael Rojas-Montes,
  • Hugo Estrada-Esquivel,
  • Axel Herroz-Herrera,
  • Jorge L. De Jesús

摘要

This work presents a new methodology for generating synthetic images. This methodology is based on a computational model of the Visual Cortex and employs multi-tree genetic programming to optimize the process. Three metrics are proposed as a fitness function to synthesize an image and are compared with each other. Likewise, the number of operations used in the process is estimated and compared with those used by a typical DCGAN, concluding that our proposal requires fewer operations.