TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Lumalabs AI

69 pointsby downbootsover 1 year ago

7 comments

mkaicover 1 year ago
The speed with which Luma transitioned their product from being NeRF-based to 3D-Gaussians-based following the release of the 3D Gaussian Splatting paper was really impressive in my opinion. It's one thing to implement a brand new research paper, it's another to implement it in a polished manner and ship it as a working product in just a few weeks!
评论 #37921739 未加载
tomohelixover 1 year ago
Am I old or is technology moving too fast these days? Or maybe before I have no idea what is possible and what isn&#x27;t so I accept incredible stuff as normal but now the &quot;impossible&quot; becomes trivial and my world view is challenged?<p>Anyway, incredible stuff. I remember the days where a 3D model took a week to make in low res. Maybe the singularity isn&#x27;t that far away...
IanCalover 1 year ago
What&#x27;s the current best open source version of this kind of thing? Awkward to get running and a UI that&#x27;s a half working notebook is fine, but are there decent models&#x2F;setups to go from video&#x2F;images to point cloud and renderable scene (with Gaussian splatting).
评论 #37922225 未加载
houseatrielahover 1 year ago
This level of fidelity will make the metaverse a compelling product.
评论 #37924633 未加载
saleemdogarover 1 year ago
Saleem
syntaxingover 1 year ago
Is it possible to 3D print an object from a 3D Gaussian based scan?
评论 #37922434 未加载
评论 #37922612 未加载
m3kw9over 1 year ago
Can do that with iPhone with LiDAR
评论 #37922574 未加载
评论 #37922578 未加载