Highlights
- Pro
real2sim
Accompanying library for the Record3D iOS app (https://record3d.app/). Allows you to receive RGBD stream from iOS devices with TrueDepth camera(s).
Grounded SAM 2: Ground and Track Anything in Videos with Grounding DINO, Florence-2 and SAM 2
[ECCV 2024] PowerPaint, a versatile image inpainting model that supports text-guided object inpainting, object removal, image outpainting and shape-guided object inpainting with only a single model…
[ECCV 2024] The official implementation of paper "BrushNet: A Plug-and-Play Image Inpainting Model with Decomposed Dual-Branch Diffusion"
Depth Pro: Sharp Monocular Metric Depth in Less Than a Second.
This code corresponds to simulation environments used as part of the DexMimicGen project.
[CVPR 2024 Highlight] FoundationPose: Unified 6D Pose Estimation and Tracking of Novel Objects
A software framework integrating various imitation learning methods and benchmark environments for robotic manipulation
GeoCalib: Learning Single-image Calibration with Geometric Optimization (ECCV 2024)
Official implementation of Crossing the Human-Robot Embodiment Gap with Sim-to-Real RL using One Human Demonstration
[CVPR 2025 Best Paper Award] VGGT: Visual Geometry Grounded Transformer
The official repository for Kalib: Markerless Hand-Eye Calibration with Keypoint Tracking.
Humanoid dataset for learning
Official implementation of Continuous 3D Perception Model with Persistent State
Code for the project "MegaSaM: Accurate, Fast and Robust Structure and Motion from Casual Dynamic Videos"
Official implementation of "DepthLab: From Partial to Complete"
Towards a Generative 3D World Engine for Embodied Intelligence
Official repo for GraspGen: A Diffusion-based Framework for 6-DOF Grasping
[ICCV 2025] InstaScene: Towards Complete 3D Instance Decomposition and Reconstruction from Cluttered Scenes
[CVPR'26] ObjectClear: Precise Object and Effect Removal with Adaptive Target-Aware Attention
[CVPR'24] MorpheuS: Neural Dynamic 360° Surface Reconstruction from Monocular RGB-D Video
3D Object Reconstruction project is a workflow that takes a set of stereo images and camera info and outputs a textured mesh (i.e., .OBJ file). The purpose is to translate physical items into the d…
Twisting Lids Off with Two Hands [CoRL 2024]