<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
  <title>3D&amp;4D NEWS｜3D &amp; 4D tech news, every 3 days</title>
  <link>https://izutsuya.io/3d4d-news/en/</link>
  <atom:link href="https://izutsuya.io/3d4d-news/en/feed.xml" rel="self" type="application/rss+xml"/>
  <description>News on 3D Gaussian Splatting, 4DGS, point clouds, volumetric video and more 3D/4D technology, published every 3 days at 7:00 JST. By IZUTSUYA Inc..</description>
  <language>en</language>
  <lastBuildDate>Mon, 02 Nov 2026 07:00:00 +0900</lastBuildDate>
  <item>
    <title>GOFish: A Method to Obtain Both a Beyond-180° 3DGS and a Collision Mesh from Fisheye Images</title>
    <link>https://izutsuya.io/3d4d-news/en/2026/11/02/01.html</link>
    <guid isPermaLink="true">https://izutsuya.io/3d4d-news/en/2026/11/02/01.html</guid>
    <category>Research</category>
    <pubDate>Mon, 02 Nov 2026 07:00:00 +0900</pubDate>
    <description>A research team including Amazon has released GOFish, a method that builds a full-field-of-view 3DGS and a well-aligned mesh from fisheye images alone. It aims to produce simulation-ready digital twins without LiDAR.</description>
  </item>
  <item>
    <title>LibreYOLO Adds Python Support for Sperildabs Idol, Which Creates 3DGS From a Single Image</title>
    <link>https://izutsuya.io/3d4d-news/en/2026/11/02/02.html</link>
    <guid isPermaLink="true">https://izutsuya.io/3d4d-news/en/2026/11/02/02.html</guid>
    <category>Tools</category>
    <pubDate>Mon, 02 Nov 2026 07:00:00 +0900</pubDate>
    <description>LibreYOLO&#39;s documentation now includes instructions for running Speridlabs Idol, which generates 3D Gaussian Splats from a single image of an object. Output is .ply and .splat, and the main license for code and weights is Apache-2.0.</description>
  </item>
  <item>
    <title>speridlabs releases &#34;Idol,&#34; a model that generates 3DGS from a single image, on Hugging Face</title>
    <link>https://izutsuya.io/3d4d-news/en/2026/11/02/03.html</link>
    <guid isPermaLink="true">https://izutsuya.io/3d4d-news/en/2026/11/02/03.html</guid>
    <category>Tools</category>
    <pubDate>Mon, 02 Nov 2026 07:00:00 +0900</pubDate>
    <description>speridlabs has released Idol, a model that generates up to about 1.05 million Gaussians from a single image of an object. Peak VRAM is 7.2GB, and the output is a standard 3DGS PLY.</description>
  </item>
  <item>
    <title>Two 3DGS Add-ons for Blender: Features and Differences Between SkySplat and the KIRI Version</title>
    <link>https://izutsuya.io/3d4d-news/en/2026/11/01/01.html</link>
    <guid isPermaLink="true">https://izutsuya.io/3d4d-news/en/2026/11/01/01.html</guid>
    <category>Tools</category>
    <pubDate>Sun, 01 Nov 2026 07:00:00 +0900</pubDate>
    <description>This article sorts out two open-source add-ons for working with Gaussian Splatting in Blender: SkySplat and 3DGS Render. The former focuses on training from drone video; the latter focuses on editing, rendering and simple lighting.</description>
  </item>
  <item>
    <title>3DGS Generates Training Data for Wrist-Camera VLA Policies, Validated on Reaching in Motherboard Disassembly</title>
    <link>https://izutsuya.io/3d4d-news/en/2026/11/01/02.html</link>
    <guid isPermaLink="true">https://izutsuya.io/3d4d-news/en/2026/11/01/02.html</guid>
    <category>Robotics</category>
    <pubDate>Sun, 01 Nov 2026 07:00:00 +0900</pubDate>
    <description>Zuoxu Wang and colleagues have posted to arXiv a method that reconstructs a real environment with 3DGS and synthesizes wrist-camera images. They report that, on reaching motions in motherboard disassembly, it achieved a higher success rate than teleoperated demonstrations.</description>
  </item>
  <item>
    <title>SG^2SLAM, a 3DGS SLAM for Monocular Outdoor Navigation, Published on arXiv</title>
    <link>https://izutsuya.io/3d4d-news/en/2026/11/01/03.html</link>
    <guid isPermaLink="true">https://izutsuya.io/3d4d-news/en/2026/11/01/03.html</guid>
    <category>Research</category>
    <pubDate>Sun, 01 Nov 2026 07:00:00 +0900</pubDate>
    <description>Ziang Guo and colleagues propose SG^2SLAM, a 3DGS SLAM for outdoor monocular navigation. They say it outperformed conventional frame-by-frame correspondence methods on KITTI and the Waymo Open Dataset.</description>
  </item>
  <item>
    <title>VinUniversity Releases Paper and Code on Deciding 3DGS Gaussian Birth and Death via the First Variation of the Loss</title>
    <link>https://izutsuya.io/3d4d-news/en/2026/11/01/04.html</link>
    <guid isPermaLink="true">https://izutsuya.io/3d4d-news/en/2026/11/01/04.html</guid>
    <category>Research</category>
    <pubDate>Sun, 01 Nov 2026 07:00:00 +0900</pubDate>
    <description>A method that controls 3DGS density using a score derived from the reconstruction loss, rather than heuristics, has been released. Code for standard 3DGS is on GitHub, and results on the seven Mip-NeRF360 scenes report higher metrics with fewer Gaussians.</description>
  </item>
  <item>
    <title>VAST-AI&#39;s Hugging Face Space &#34;TripoSplat&#34; Generates a 3DGS PLY from a Single Image</title>
    <link>https://izutsuya.io/3d4d-news/en/2026/11/01/05.html</link>
    <guid isPermaLink="true">https://izutsuya.io/3d4d-news/en/2026/11/01/05.html</guid>
    <category>Tools</category>
    <pubDate>Sun, 01 Nov 2026 07:00:00 +0900</pubDate>
    <description>VAST-AI has released &#34;TripoSplat,&#34; a demo on Hugging Face that converts images into Gaussian splats. A post says the output PLY can be opened in SuperSplat, a browser-based editing tool.</description>
  </item>
</channel>
</rss>