The undisputed king of micro splatting @DanyBittel has generously shared a bunch of his data. Go check it out now, it's really an amazing gift! We threw it into KIRI Engine to check how our 3DGS to Mesh 3.0 would perform on a micro level and as hoped, good data in = good data out 😆

I don't know what I expected but Linux gaming just works at this point. I'm on a fresh install of Fedora 44 on an NVIDIA GPU and everything just works. What's going on? Is 2026 really the year of Linux?

After years of trying, giving up, and starting again… I made it into the @YouTube Partner Program!🥹 Thanks a lot if you watched, liked, or stayed even a few seconds, I truly appreciate ur support! 🤍 We're just getting started!🔥 🔗👇 youtube.com/@MadMaraca

Quest Browser 146.o now ships with experimental support for WebGPU in #WebXR. Turn on "webXR experimental features" in "chrome://flags" and the API will be available. Feedback welcome!

Today, we released Lyra 2.0, a framework for generating persistent, explorable 3D worlds at scale, from NVIDIA Research. Generating large-scale, complex environments is difficult for AI models. Current models often “forget” what spaces look like and lose track of movement over time, causing objects to shift, blur, or appear inconsistent. This prevents them from creating the reliable 3D environments required for downstream simulations. Lyra 2.0 solves these issues by: ✅ Maintaining per-frame 3D geometry to retrieve past frames and establish spatial correspondences ✅ Using self-augmented training to correct its own temporal drifting. Lyra 2.0 turns an image into a 3D world you can walk through, look back, and drop a robot into for real-time rendering, simulation, and immersive applications. ➡️ Learn more: research.nvidia.com/labs/sil… 📄 Read the paper: arxiv.org/abs/2604.13036

Spark 2.0 Load humongous splats with the new LoD system. Enjoy!

Spark 2.0 is here! 🚀 We’re redefining what’s possible on the web with a streamable LoD system for 3D Gaussian Splatting. Built on Three.js, you can now stream massive 100M splat worlds to any device from mobile to VR using WebGL2. All open-source. Dive into the tech 👇

Heads up: there are some changes coming to LichtFeld Studio. LichtFeld is growing quite a lot, but contributions are not. In fact, besides the two corporate contributors and one bronze sponsor, there is less than $100 on average contributed per month. There were 10,000 downloads in the last 30 days, and nobody contributed anything financially back to the project. Not a single dollar. I cannot support my living on that basis. Given the amount of work I am putting into this project, and it is a lot, often easily 60 hours a week, to provide a state-of-the-art, self-contained 3D reconstruction suite that is used by businesses around the world, this situation is no longer sustainable. The idea was to grow it together, but at the moment it is mainly being used without enough support in return. If every user had contributed only $10, the continuation of the project would have been secured. I have asked for that quite often. I like sharing my knowledge, but I cannot sustain it under these circumstances. Going forward, there will not be any free binaries provided anymore. I created a portal at portal.lichtfeld.io which will give access to daily binaries after registration and donation. This will hopefully help finance the development and maybe even allow me to pay someone in addition. If you cannot afford it, you can still put in the work and build it from source. There are no restrictions. I am aware that people like to work around something like this, and they certainly will. But in the end, that simply means that my work on this project will be discontinued at some point. Btw, after payment an invoice will be automatically issued!

The XUASTC image/texture format is just the beginning: it's essentially 1992 JPEG-style DCT ported into ASTC (weight grid DCT). The next big step will be porting WebP's intra-prediction into the weight grid domain. The result will be much better quality per bit at low bitrates.

Gemma 4 can run on phones without an internet connection! 🤯 It can perform local agentic tasks, such as logging and analyzing trends. When connected, it can also make API calls. Want to try it yourself? Get the Google AI Edge App on iOS or Android. (🔊 Sound on for the demo!)

🚀 Excited to release the code and models of ReSplat: Learning Recurrent Gaussian Splatting! 🌐 Project page: haofeixu.github.io/resplat/ 💻 Code & models: github.com/cvg/resplat

what if three.js scenes had a real DSP brain? replaced basic Web Audio nodes w/ a Faust sound engine: DSP → WASM → browser codex & i built a custom web ui control surface to play & tune drone fx dsp params before mapping them to three.js visuals next post i'll share full interactive #threejs experiment #faust #webaudio #dsp

You can now render humongous worlds! Enjoy and let us know what you think!

vLLM v0.15.0 project update is out. In this 15-minute update, @mgoin_ shares: 🆕 What’s new in vLLM v0.15.0 🪲 Debugging a real memory leak in vLLM 🧩 Mixture-of-Models on AMD GPUs with vLLM-SR 🧠 New model support: DeepSeek-OCR 2, Kimi K2.5, Arcee Trinity-Large, Qwen3-ASR

CuMesh: High-Performance Geometry Processing for PyTorch Cuda mesh utils. github.com/JeffreyXiang/CuMe… CuMesh is a GPU-accelerated library designed for high-performance 3D geometry processing directly within the PyTorch ecosystem. It provides efficient primitives for mesh cleaning, decimation, remeshing, and UV unwrapping. Key features include: ・CUDA-Accelerated Mesh Operations: Fast topology queries, simplification, hole filling, and cleaning. ・Remeshing: Remesh arbitrary meshes using narrow-band UDF and Dual Contouring. ・UV Unwrapping: Efficient UV parameterization with xatlas enhenced by a fast mesh clustering on the GPU.

I defined a Three.js vendor extension to serialize Three.js TSL into glTF and deserialize it from glTF, and I tried creating GLTFLoader/Exporter plugins for Three.js to handle that extension. github.com/takahirox/THREE_m… #threejs #tsl #gltf #vibecoding

Basis Universal v2.0, our GPU supercompressed texture library, is now live on GitHub. This update adds full ASTC LDR and supercompressed XUASTC LDR (supporting all 14 ASTC block sizes) with Weight Grid DCT, and new real-time Analytical BC7/ETC1 encoders. github.com/BinomialLLC/basis…

With the help @aframevr I dusted off my old daylight system component and turned it into a nice 360 32-bit panorama exporter. Sun size, 8k, artificial lights, night sky and other cool updates are on the way! Start making you own hdri panos here! ex3d.com/hdri @kfarr

I've been working a lot with SAM3 and the Momentum Human Rig (MHR). I finally integrated it into the data I'm working with @rerundotio. The progression I've taken looks as follows SAM3 SAM3D-body on 1. a single image 2. a set of multiple images 3. a single video 4. A multiview video capture I took inspiration from the SAM3D-body paper and built a multiview fitting optimization pipeline. This pipeline involves using the 2D keypoints from the single-view pipeline, triangulating them, and employing an L1 loss between the 2D/3D keypoints. The temporal stability isn't great, so that's the next portion I'm going to focus on. One really frustrating thing about SAM3D-body is the lack of per-joint confidence values. It makes it harder to deal with occlusions. I'm probably going to need to use a separate model, or maybe add a confidence head.

JoltPhysics.js v1.0.0 is out github.com/jrouwe/JoltPhysic…