⛳️Happy to present SCube, our newest 3D large-scale scene reconstruction model. It has been accepted by #NeurIPS2024! Great job Xuanchi Ren and Yifan!

📢Please check out our newest work on feed-forward reconstruction of dynamic monocular videos! With our bullet-time formulation, we reach great flexibility and state-of-the-art performance!

New #NVIDIA paper to make diffusion models better and faster 🚀 Multi-Student Distillation! We distill diffusion models into multiple 1-step students, allowing (a) improved quality by specializing in subsets and (b) improved latency by distilling into smaller architectures. 1/n

🚀Excited to Introduce #BTimer: Real-Time Dynamic Scene Reconstruction from Monocular Videos! Struggling with novel view synthesis on dynamic scenes? Meet BTimer (BulletTimer) — the 1st motion-aware feed-forward model for real-time scene reconstruction at any desired time. ✅

Built upon X/SCube, our newest InfiniCube generates controllable dynamic 3DGS driving scenes from an HD map and bounding boxes. Generating these simulation environments unlocks many more possibilities in autonomous driving! Kudos to Yifan Lu and Xuanchi Ren! 🥳

Introducing InfiniCube, a scalable method for generating unbounded dynamic 3D driving scenes with high fidelity and controllability.

Spatio-Temporal Reconstruction Model for Large-Scale Outdoor Scenes Contributions: • We propose STORM, the first feed-forward, self-supervised method for fast and accurate reconstruction of dynamic 3D scenes from sparse, multi-timestep, posed camera images. • Our bottom-up

Whether you're a researcher or developer, #NVIDIACosmos world foundation models are now openly available under our permissive license to the physical AI community via NGC & Hugging Face. 🤗 See how Cosmos is democratizing #physicalAI development. #CES2025 bit.ly/4g4tVkR

MATCHA:Towards Matching Anything feixue, Sven Elflein, Laura Leal-Taixe, Qunjie Zhou tl;dr: diffusion model->semantic+geometric features->transformer-based fusion->enhanced diffusion features->w/ DINOv2->unified feature->geometric/semantic/temporal matching arxiv.org/abs/2501.14945

Excited to share our #CVPR2025 paper: Difix3D+ Difix3D+ reimagines 3D reconstruction with single-step diffusion, distilling 2D generative priors for realistic novel view synthesis from large viewpoint shifts. 📄Paper: arxiv.org/abs/2503.01774 🌐Website: research.nvidia.com/labs/toronto-a…

📽️🔥 Please checkout GEN3C, a video model with an explicit 3D cache inside for better consistency and precise camera control! #CVPR2025 #NVIDIA

📢 GEN3C is now open-sourced, with code released under Apache 2.0 and model weights under the NVIDIA Open Model License! 🚀 Along with it, we're releasing a GUI tool that lets you specify your desired video trajectory in 3D — come play with it and generate your own! The

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,