<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Text-to-3d on Ryan Orban</title><link>https://ryanorban.com/categories/text-to-3d/</link><description>Recent content in Text-to-3d on Ryan Orban</description><generator>Hugo</generator><language>en-us</language><managingEditor>me@ryanorban.com (Ryan Orban)</managingEditor><webMaster>me@ryanorban.com (Ryan Orban)</webMaster><copyright>Ryan Orban</copyright><lastBuildDate>Tue, 22 Nov 2022 00:00:00 +0000</lastBuildDate><atom:link href="https://ryanorban.com/categories/text-to-3d/index.xml" rel="self" type="application/rss+xml"/><item><title>Magic3D — High-Resolution Text-to-3D Content Creation</title><link>https://ryanorban.com/notes/magic3d-text-to-3d/</link><pubDate>Tue, 22 Nov 2022 00:00:00 +0000</pubDate><author>me@ryanorban.com (Ryan Orban)</author><guid>https://ryanorban.com/notes/magic3d-text-to-3d/</guid><description>&lt;h3 id="summary" class="scroll-mt-8 group"&gt;
 Summary
 
 &lt;a href="#summary"
 class="no-underline hidden opacity-50 hover:opacity-100 !text-inherit group-hover:inline-block"
 aria-hidden="true" title="Link to this heading" tabindex="-1"&gt;
 &lt;svg
 xmlns="http://www.w3.org/2000/svg"
 width="16"
 height="16"
 fill="none"
 stroke="currentColor"
 stroke-linecap="round"
 stroke-linejoin="round"
 stroke-width="2"
 class="lucide lucide-link w-4 h-4 block"
 viewBox="0 0 24 24"
&gt;
 &lt;path d="M10 13a5 5 0 0 0 7.54.54l3-3a5 5 0 0 0-7.07-7.07l-1.72 1.71" /&gt;
 &lt;path d="M14 11a5 5 0 0 0-7.54-.54l-3 3a5 5 0 0 0 7.07 7.07l1.71-1.71" /&gt;
&lt;/svg&gt;

 &lt;/a&gt;
 
&lt;/h3&gt;
&lt;p&gt;Magic3D is a text-to-3D generation system from NVIDIA Research that creates high-resolution textured 3D meshes from text prompts. The method is notable for its two-stage pipeline: first it generates a coarse NeRF representation using DreamFusion&amp;rsquo;s score distillation sampling (SDS) approach, then refines it into a high-resolution mesh with better surface detail and texture. This two-stage coarse-to-fine approach was faster than pure NeRF optimization and produced cleaner meshes for downstream use.&lt;/p&gt;</description></item></channel></rss>