<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>大模型影像论文精读 on 花花世界</title>
    <link>https://shuweiyue.com/ai/llm-imaging/</link>
    <description>Recent content in 大模型影像论文精读 on 花花世界</description>
    <generator>Hugo</generator>
    <language>zh-cn</language>
    <lastBuildDate>Tue, 07 Jul 2026 00:00:00 +0800</lastBuildDate>
    <atom:link href="https://shuweiyue.com/ai/llm-imaging/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>让大模型先看一眼颜色对不对：VLM-CC 把色彩恒常做成了反馈闭环</title>
      <link>https://shuweiyue.com/ai/llm-imaging/2026-07-07-vlm-cc-cross-camera-color-constancy/</link>
      <pubDate>Tue, 07 Jul 2026 00:00:00 +0800</pubDate>
      <guid>https://shuweiyue.com/ai/llm-imaging/2026-07-07-vlm-cc-cross-camera-color-constancy/</guid>
      <description>VLM-CC 不让大模型直接回归光源颜色，而是先做一次白平衡，再让 VLM 判断画面还偏红、偏绿还是偏蓝，最后把这个语义反馈转成下一轮照明估计。它的价值在于把大模型放到颜色还原的评审环节，而不是粗暴替代传统 ISP。</description>
    </item>
  </channel>
</rss>
