<?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>Physical AI on rocPAI-Forge</title><link>https://rocpai-forge.github.io/en/tags/physical-ai/</link><description>Recent content in Physical AI on rocPAI-Forge</description><generator>Hugo</generator><language>en</language><lastBuildDate>Thu, 10 Sep 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://rocpai-forge.github.io/en/tags/physical-ai/index.xml" rel="self" type="application/rss+xml"/><item><title>Helping Physical AI Understand Dynamic Worlds: Building 4DGS from Monocular Video on AMD ROCm</title><link>https://rocpai-forge.github.io/en/posts/amd-4dgs-series-01/</link><pubDate>Thu, 10 Sep 2026 00:00:00 +0000</pubDate><guid>https://rocpai-forge.github.io/en/posts/amd-4dgs-series-01/</guid><description>&lt;h2 id="a-dynamic-world-representation-for-physical-ai">A dynamic-world representation for Physical AI&lt;/h2>
&lt;p>Physical AI is not only about generating pixels; it is about sensing, reconstructing, and querying a changing real world. Conventional 4DGS commonly starts from synchronized, calibrated multi-view video. This article explores a lower-capture-barrier path: start from monocular video, synthesize synchronized multi-view observations with a 4D video-generation model, and build a queryable 4DGS representation of the dynamic world.&lt;/p>
&lt;p>Within rocPAI-Forge’s Physical AI direction, this pipeline is a &lt;strong>dynamic visual layer&lt;/strong>:&lt;/p></description></item></channel></rss>