<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>🎯 目标引导 on Elon&#39;s AD Insight</title>
    <link>https://auto-driving-blog.vercel.app/tags/-%E7%9B%AE%E6%A0%87%E5%BC%95%E5%AF%BC/</link>
    <description>Recent content in 🎯 目标引导 on Elon&#39;s AD Insight</description>
    <generator>Hugo</generator>
    <language>zh-cn</language>
    <lastBuildDate>Sun, 19 Jul 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://auto-driving-blog.vercel.app/tags/-%E7%9B%AE%E6%A0%87%E5%BC%95%E5%AF%BC/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>论文精读｜GoalFlow：目标引导流匹配轨迹生成</title>
      <link>https://auto-driving-blog.vercel.app/posts/paper-reading/goalflow%E7%B2%BE%E8%AF%BB/</link>
      <pubDate>Sun, 19 Jul 2026 00:00:00 +0000</pubDate>
      <guid>https://auto-driving-blog.vercel.app/posts/paper-reading/goalflow%E7%B2%BE%E8%AF%BB/</guid>
      <description>GoalFlow 用一个精准目标点词表加评分机制选出最合理的短期终点，再用 Flow Matching 一步生成多模态驾驶轨迹。它解决了纯扩散生成轨迹发散无边界、而固定锚点约束过强不真实的两难问题。在 NAVSIM 上 PDMS 达到 90.3，验证了&amp;rsquo;目标点+整流流&amp;rsquo;高效生成方案的有效性。</description>
    </item>
  </channel>
</rss>
