Simulation drives robotics progress, but how do we close the reality gap?
Introducing GaussGym: an open-source framework for learning locomotion from pixels with ultra-fast parallelized photorealistic rendering across >4,000 iPhone, GrandTour, ARKit, and Veo scenes!
Thread 🧵
Learn how the Radiance Field method, Gaussian Splatting gets compressed with @wimost and @FlorianIBarthel's Self Organizing Gaussians.
Self Organizing Gaussians is also what @nerfstudioteam uses for gsplat.
Full episode is now live here:
piped.video/Y0O6R0Keywg
Ready to take your 3DGS scene capturing skills to the next level?
Check out this EPIC video tutorial @simonbethke just dropped!
Learn how to capture stunning scenes with step-by-step explanations and discover some insider camera hacks that will instantly boost your results.
He also is breaking down how to create the perfect sparse point cloud and train your scene like a pro!
If you’re into radiance fields or just love pushing the boundaries of what’s possible with 3D scenes, this one’s for you
This is a sneak peak. Please find the link to the full tutorial in the comments.
Check out @meta_aria data being used to create radiance fields of my back yard from a pair of glasses! We've added full support for Project Aria data into Nerfstudio and Gsplat. Thanks @nerfstudioteam for the help!
The following NeRF was created in Nerfstudio using egocentric data from Project Aria glasses. 👀
Did you know that Project Aria data is supported in #Nerfstudio? 🚀
Try it out yourself using the Nerfstudio docs: docs.nerf.studio/quickstart/…
3D Gaussians Splats look good, but they are less interesting when static.
Our ECCV 2024 work, Feature Splatting, enhances 3DGS with foundational vision features and physical simulation to synthesize dynamic scenes from static 3D captures.
✨Open-sourced in NerfStudio!👇
🌟gsplat🌟 (docs.gsplat.studio) now supports multi-GPU distributed training, which nearly linearly reduces the training time and memory footprint.
Now Gaussian Splatting is ready for the city-scale reconstruction!
kudos to this amazing paper: daohanlu.github.io/scaling-u…
Wrapping up this incomplete and unrepresentative hacky poll: Nerfstudio seems to dominate the splats, followed by Luma.
Some also mention Scaniverse as a popular app, even though it was not part of the poll.
However, there are many different options to choose from. Awesome!
Excited to release 🚀Nerfstudio Gradio WebUI🚀: Simplify your 3D model training & visualization with an intuitive GUI!
github.com/nerfstudio-projec…
📊 Train models directly from your browser
👁️ Visualize with Viser
📦 Process data seamlessly
🚀 Export models with ease
And more...
Excited to announce 🚀gsplat v1.0🚀: a ⏩efficient⏩ CUDA backend for 3D Gaussian Splatting!
docs.gsplat.studio/
A drop-in replacement of the official impl. with:
- Up to 2x faster training;
- Up to 4x less GPU memory;
- Render millions of GSs in real-time;
- And more;
While we love yellow bulldozers, yellow subsea robots are also just as swaggy. Check out this NeRF made by @ScottyDempsey82 using Nerfstudio's nerfacto-big! piped.video/watch?v=ZbGENrLT…
SigNeRF from @JDihlmann, @AjlEngelhardt
& @CG_Tuebingen introduces ControlNet to NeRFs, offering control and easy manipulation in NeRF inpainting. SigNeRF is modular and run on standard Nerfacto, so it should be translatable to Gaussian Splatting!
🔗radiancefields.com/controlne…
Filmed sakura pre-blooming in Kyoto last week with the A7Siii
Testing out processing into Gaussian Splats using @nerfstudioteam (thanks for the tutorial @janusch_patas !) and @jawset
Real-time rendering and transformations between the two outputs at 4K 60fps in @unity
watch on YT
🎉 V1.0.0 officially released!
* New splatfacto model, nerfstudio's implementation of Gaussian splatting
* Complete rewrite of the viewer, now built entirely on viser
If you've been using Nerfstudio and have any feedback for us, please consider filling out our feedback form @ forms.gle/sqN5phJN7LfQVwnP9. Thank you! 🥳
...aaaand here is an update to todays flower scan.
This time using @nerfstudioteam to create the pointcloud. The vase is a little wobbly due to it's reflective nature, but the definition of the leaves, stems and petals is impressive.