Scholarlink

<p>This study investigates the application of artificial intelligence in automated painting creation and human–machine collaborative innovation. To improve the emotional expression and creative flexibility of AI-generated paintings, an Emotional Semantic Guided Generation Model (ESGGM) was proposed. The model integrates emotional semantic encoding, style transfer and fusion, and creative intention reasoning. In addition, a human–machine collaborative creation system was designed to support user input, bidirectional feedback and personalized model adaptation. Experimental results show that ESGGM achieved a comprehensive artistic value score of 0.82, outperforming the CNN style transfer model, GAN and NST. The proposed collaborative system achieved a score of 0.85, higher than the rule-based interaction model and the sketch-to-painting model. These results indicate that emotional semantic guidance and interactive collaboration can improve the quality and controllability of AI-assisted painting creation.</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

Application of artificial intelligence in automated painting creation and human–machine collaborative innovation

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月28日

查看全文
  • Xu Xu,
  • Manqiu Xu,
  • Yongsheng Zhang

摘要

This study investigates the application of artificial intelligence in automated painting creation and human–machine collaborative innovation. To improve the emotional expression and creative flexibility of AI-generated paintings, an Emotional Semantic Guided Generation Model (ESGGM) was proposed. The model integrates emotional semantic encoding, style transfer and fusion, and creative intention reasoning. In addition, a human–machine collaborative creation system was designed to support user input, bidirectional feedback and personalized model adaptation. Experimental results show that ESGGM achieved a comprehensive artistic value score of 0.82, outperforming the CNN style transfer model, GAN and NST. The proposed collaborative system achieved a score of 0.85, higher than the rule-based interaction model and the sketch-to-painting model. These results indicate that emotional semantic guidance and interactive collaboration can improve the quality and controllability of AI-assisted painting creation.