Open source Structure-from-Motion pipeline

CoTracker is a model for tracking any point (pixel) on a video.

High-resolution models for human tasks.

Detectron2 is a platform for object detection, segmentation and other visual recognition tasks.

High-Resolution 3D Human Digitization from A Single Image.

Pippo: High-Resolution Multi-View Humans from a Single Image

This repository provides inference code to compute canopy height maps from aerial images, as described in the paper "Very high resolution canopy height maps from RGB imagery using self-supervised v…

Code release for "Cut and Learn for Unsupervised Object Detection and Instance Segmentation" and "VideoCutLER: Surprisingly Simple Unsupervised Video Instance Segmentation"

[CVPR'25 Highlight] Official repository of Sonata: Self-Supervised Learning of Reliable Point Representations

Segment This Thing is an efficient image segmentation models that uses a biologically-inspired foveated tokenization to reduce inference costs.

[CVPR 2025 Best Paper Award] VGGT: Visual Geometry Grounded Transformer

Official implementation of the ICCV25 paper "Easy3D A Simple Yet Effective Method for 3D Interactive Segmentation".

Open source impl of **MV-DUSt3R+ Single-Stage Scene Reconstruction from Sparse Views In 2 Seconds** from Meta Reality Labs. Project page https://mv-dust3rp.github.io/

A high-performance C++ header and application for real-time metric depth estimation, using models from Depth-Anything-V2. Powered by ONNX Runtime and OpenCV, it supports seamless inference for imag…

A high-performance C++ headers for real-time object detection and segmentation using YOLO models, leveraging ONNX Runtime and OpenCV for seamless integration. Supports multiple YOLO (v5, v7, v8, v9…

C++ TensorRT implementation of Depth-Anything V1, V2

SwinIR: Image Restoration Using Swin Transformer (official repository)

A simple webui for stable-diffusion.cpp

Model deployment based TensorRT

Tools for the Embody 3D Dataset

Official SeedVR2 Video Upscaler for ComfyUI

Nodes for using ComfyUI as a backend for external tools. Send and receive images directly without filesystem upload/download.

Computer Vision ML inference in C++

A graph node engine and editor written in Javascript similar to PD or UDK Blueprints, comes with its own editor in HTML5 Canvas2D. The engine can run client side or server side using Node. It allow…

Tensor library for machine learning

Detect Anything via Next Point Prediction (Based on Qwen2.5-VL-3B)

Declarative User Interfaces for Python

An open-source, GPU-accelerated physics simulation engine built upon NVIDIA Warp, specifically targeting roboticists and simulation researchers.

Gaia Physics Engine