XGRIDS World Backup Project でグランプリをいただきました🏆
3つの水辺を歩ける3Dインタラクティブ体験です。
Honored to receive the Grand Prize at #WorldBackupProject — thank you, XGRIDS!
🔗 https://t.co/CiuAM7TUQu
At the customer’s option, our team is exploring the possibility of integrating every truck with LoRA mesh capability.
This technology would allow REO owners to communicate across the country for free over encrypted networks.
We’re considering this for disaster scenarios and to put communication in the hands of the people.
The BIS Connected Vehicles rule prohibits communication devices in automobiles with components from China so we will be building our own.
Remember, every Runabout will ship in the most analog state possible so our customers can choose how much technology they want to adapt.
The world's largest 3DGS town reconstruction! 🌍
23.5 km². 6.2 billion Gaussian Splats.
Captured with #LixelL2Pro & #PortalCam, this entire town becomes a virtual film set in Unreal Engine, powered by LCC plugin for UE
👉 https://t.co/P2TTuETd93
What do you do if you have a few hours ashore on a cruise to Puerto Rico? 🛳️
Scan historic San Juan and upload it to SuperSplat!
1.39GB of Gaussian splats, streamed and rendered on WebGPU by the @playcanvas engine.
[1/2]
🤖 Explore AI in the classroom with our Ambassador-created AI Buddy Cultural Classroom lesson plan! Students design a 3D cultural environment and create an AI Buddy language tutor. 🌎
Download lesson plan 👉 https://t.co/0DelfJMOwq
FlyOS-Fast is a lightweight Linux system optimized for 3D printing, developed by Mellow. It provides FLY series host boards and all-in-one mainboards with a unified environment for print control, web-based management, local display, camera support, and system updates.
#3dprint
Check out my Space on @arrival_space_
I made this in @theworldlabs#Marble (with a 360 image and a prompt) It's a one room test but I'll be expanding this to a multi-room space in the near.
https://t.co/0L34fgwQ0L #ArrivalSpace
The result of reprocessing an old OSMO360 data with the latest my workflow (Dual SAM3 + Seam mask + Cubemap reduction). See in 4K 😀
(The lower shrine in previous post leads to the upper shrine on the summit in Oyama)
---Steps---
1. DJI OSMO360 10min walk ("RockSteady off")
2. Frame extraction and masking in my 3DGS tool : 851images
3. Spherical SfM in Metashape std. with the masks
4. Cubemap conversion with the mask and cubemap reduction (17% cutoff) in my 3DGS tool
5. 3DGS train by LFS, RTX 4090, 15M, SH2, PPISP
6. Cam work by LFS
-----------
Of course, my tools and workflow are available for Insta360 X6. If you try longer 3DGS than 2 min limitation of Insta360 3DGS, please let's try this(See comment).😊
💧 121 million Gaussians
📡 1.3GB of Streamed SOG data
🧊 11 million collision voxels
This scan of Vyšehrad is one of the largest 3D Gaussian splats ever. SuperSplat delivers it to your browser and @playcanvas renders it on WebGPU.
So big, I had to speed up the video 2.5x!
[1/2]
Full #3DGS of an apartment made in @UnrealEngine , Gaussian Splat trained with @lichtfeldstudio
with 5M splats and 10k steps. #gaussiansplats#unrealengine
Starting to get some good fidelity on complex materials. Although some spaces need more data, we are getting close.
How accurate is our automatic 360° reconstruction pipeline?
We tested it against a LIDAR survey captured at the same time, on the same rig, over a 118m path.
The result: 3.9 cm median agreement with LIDAR, with no measurable accumulated drift across the full trajectory.
The one systematic discrepancy we found was a uniform 2.9% scale offset. Our reconstruction reproduces its own metric reference to 0.03%, showing that the geometry itself did not create the error.
In other words, the hard part, reconstructing shape consistently over a large scene, is already working at centimeter-level agreement.
Read the full OVER Research publication:
https://t.co/tVyskKyrlu
Okay, Meshy 7 isn’t messing around.
I tested their latest image to 3D model and I’m seriously impressed. The detail it created from a single image is wild.