[ICCV'25 Best Paper Finalist] ReCamMaster: Camera-Controlled Generative Rendering from A Single Video

ARTDECO unifies 3D foundation priors with structured scene representations, enabling robust and generalizable 3D reconstruction of diverse real-world scenes using only monocular video.

WorldPlay: Interactive World Modeling with Real-Time Latency and Geometric Consistency

RealSee3D: A multi-view RGB-D dataset combining real-world captures and procedurally generated scenes, with extensible annotations for diverse 3D vision research.

Fast-FoundationStereo: Real-Time Zero-Shot Stereo Matching

Learning Dexterous Manipulation Skills from Imperfect Simulations

G2VLM: Geometry Grounded Vision Language Model with Unified 3D Reconstruction and Spatial Reasoning

A list of works on video generation towards world model

This is the official implementation of our paper: "MiniMax-Remover: Taming Bad Noise Helps Video Object Removal"

Offical codes for "AutoVFX: Physically Realistic Video Editing from Natural Language Instructions."

Any4D: Unified Feed-Forward Metric 4D Reconstruction

Ego-R1: Chain-of-Tool-Thought for Ultra-Long Egocentric Video Reasoning

[CVPR 2025] EgoLife: Towards Egocentric Life Assistant

Python wrapper for the NVIDIA cuSFM library

[NeurIPS 25] TrackingWorld: World-centric Monocular 3D Tracking of Almost All Pixels

A novel 4D reconstruction method that directly generates high-quality, animation-ready 4D mesh asset (.GLB file) from a single monocular video.

Official repository for the project "TraceGen: World Modeling in 3D Trace-Space Enables Learning from Cross-Embodiment Videos"

[NeurIPS 2025]"DynamicVerse: A Physically-Aware Multimodal Framework for 4D World Modeling"

The official repository of "Astra : General Interactive World Model with Autoregressive Denoising"

[NeurIPS 2025] Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance

[3DV 2026 Oral] GarmentCrafter: Progressive Novel View Synthesis for Single-View 3D Garment Reconstruction and Editing

Visionary: The World Model Carrier Built on WebGPU-Powered Gaussian Splatting Platform

cowsay is a configurable talking cow

LW-BenchHub is a unified benchmark hub built on Isaac Lab–Arena for embodied AI, providing consistent interfaces, realistic environments, multi-robot support, and large-scale evaluation. It include…

Isaac Lab - Arena is a robotics simulation framework that enhances NVIDIA Isaac Lab by providing a composable, scalable system for creating diverse simulation environments and evaluating robot lear…