Qi Wu

Say goodbye to perfect pinhole assumptions Excited to introduce 3DGUT—a Gaussian Splatting formulation that unlocks support for distorted cameras, including time dependent effects like rolling shutter, while maintaining the benefits of rasterization, rendering at >250 FPS. 🧵

🚀We just released Asset Harvester, an image-to-3D model and end-to-end pipeline that extracts real object assets from autonomous driving videos! 🌐 Website: research.nvidia.com/labs/sil/proje… 💻 Code: github.com/nvidia/asset-h… [1/5] #AssetHarvester #AVSimulation #WorldModel #AutonomousDriving

Feed-forward 3D reconstruction should not be limited to predicting one Gaussian per pixel. We introduce TokenGS, which uses learnable tokens to decouple the 3D Gaussian prediction from the image resolution and the number of input views. #CVPR2026Highlight [1/6]

A canonical open-source data platform for multi-sensor neural 3D reconstruction is here. Introducing NVIDIA NCore — a unified data format and APIs for cameras, lidars, radars, poses, calibrations, and labels, built for AV, robotics, and physical AI. 🔗research.nvidia.com/labs/sil/proje…

One of the coolest project I've been involved in NVIDIA. The quality is insane. 👀

Introducing Neural Harmonic Textures: our new method for real-time novel view synthesis that outperforms all 3DGS and NeRF derivatives including (finally) ZipNeRF in terms of quality across all benchmarks. The code is released (Apache 2.0): (research.nvidia.com/labs/sil/proje…) 🧵

This week at @NVIDIAGTC we're presenting AlpaDreams: a generative world model for driving simulation. Compared to standard video models, AlpaDreams is autoregressive, enabling updating the conditioning (simple bounding box world) in closed loop, and multiview-consistent.

Such a lovely team to work with—so many talented and devoted people. At NVIDIA, we work hard not out of fear on being fired, but because we truly enjoy the team and the project. Drop me or anyone on the team a message if you’re interested in joining. research.nvidia.com/labs/sil/membe…

Special moment to see something I’ve worked on so closely come to life! Today we announce Alpadreams — a world model that lets you explore ♾endlessly♾️in ⚡real time⚡. Video: me (left) and Alpamayo policy (right) driving in Alpadreams at #GTC26. research.nvidia.com/labs/sil/proje…

A new generation in AV simulation is here! We are announcing AlpaDreams, a real time interactive generative world model for AV simualtion! Just a year ago it took minutes to generate a few seconds of video, today it is real time and interactive! research.nvidia.com/labs/sil/proje…

One for the tinfoil hats. If this ends Verstappen, Piastri, Norris, then the Monza swap WON Norris the title over Verstappen ✍️ @natesaundersF1

Our team at Nvidia Spatial Intelligence Lab is hiring PhD research interns for 2026! research.nvidia.com/labs/sil/ If you’re excited about fast video models, generative world simulators, or 3D foundation models, please reach out by email or apply directly lnkd.in/gGKU_sUr

Nvidia just released Lyra on Hugging Face Generative 3D Scene Reconstruction via Video Diffusion Model Self-Distillation TL;DR: Feed-forward 3D and 4D scene generation from a single image/video trained with synthetic data generated by a camera-controlled video diffusion model

Viser completely changed the way we do research. Before viser, it was hard to visualize 3D/4D data, let alone share it. Now it’s all just in a browser! It’s amazingly powerful and looks awesome. It’s how we render our results and videos. We love it and hope you will too!

[1/N] 🎥 We've made available a powerful spatial AI tool named ViPE: Video Pose Engine, to recover camera motion, intrinsics, and dense metric depth from casual videos! Running at 3–5 FPS, ViPE handles cinematic shots, dashcams, and even 360° panoramas. 🔗 research.nvidia.com/labs/toronto-a…

ViPE: Video Pose Engine for 3D Geometric Perception Contributions: • A robust and efficient framework, ViPE, for estimating camera parameters and dense depth from diverse, in-the-wild videos. • A system design that integrates the strengths of classical SLAM (efficiency, scalability) and learned models (robustness), with key improvements in efficiency, dynamic object handling, and depth quality over prior work. • A large-scale dataset of annotated videos, created using ViPE, to facilitate future research in 3D computer vision.

GSCache: Real-Time Radiance Caching for Volume Path Tracing using 3D Gaussian Splatting Contributions: • We introduce a novel radiance cache optimized for volume rendering that caches path-space radiance using multiple levels of Gaussian splats. • The cache works in real time on complex datasets and in a wide variety of use cases. It adapts quickly to changes in the transfer function and lighting parameters, improving overall image quality and rendering times. • Optimizing the cache is possible not only with clean samples but also with noisy data, as is commonly found in Monte-Carlo-based renderers. • The path-space nature of the cache and its non-invasive design make it easy to use and integrate into existing rendering solutions.

Super cool!

For everyone interested in precise 📷camera control 📷 in transformers [e.g., video / world model etc] Stop settling for Plücker raymaps -- use camera-aware relative PE in your attention layers, like RoPE (for LLMs) but for cameras! Paper & code: liruilong.cn/prope/

Getting mem-bound kernels to speed-of-light isn't a dark art, it's just about getting the a couple of details right. We wrote a tutorial on how to do this, with code you can directly use. Thanks to the new CuTe-DSL, we can hit speed-of-light without a single line of CUDA C++.

We are excited to share Cosmos-Drive-Dreams 🚀 A bold new synthetic data generation (SDG) pipeline powered by world foundation models—designed to synthesize rich, challenging driving scenarios at scale. Models, Code, Dataset, Tookit are released. Website: research.nvidia.com/labs/toronto-a…