# unrealcv/unrealcv

UnrealCV: Connecting Computer Vision to Unreal Engine

Repository: https://github.com/unrealcv/unrealcv
Canonical: https://ross.abutalabs.com/products/unrealcv
Homepage: https://unrealcv.org
Language: C++
License: MIT
License Family: permissive
Topics: virtual-worlds, computer-vision, ue4, embodied-ai, machine-learning, simulation, synthetic-data, ue5
Last push: 2026-08-25T19:37:37+00:00

## Health v2 (maintenance only)
Score: 67/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 8, longevity 100
- inputs: {"age_days": 3646, "days_push": 8, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2209, forks 465 (observed 2026-08-28T04:06:26.212765+00:00)

## What it is
UnrealCV is an open-source Unreal Engine plugin that connects computer vision research to virtual worlds by exposing a command API and Python client for camera control, object manipulation, and ground-truth image capture. It supports UE4/UE5, RPC communication with PyTorch/TensorFlow, and ships with 100+ ready-made UnrealZoo environments for embodied AI.

## Use cases
- generate synthetic image datasets with ground truth annotations
- train embodied AI agents in virtual environments
- test object detection models like Faster-RCNN on rendered images
- capture optical flow and depth from Unreal Engine scenes
- control Unreal Engine cameras and objects from Python
- build virtual worlds for computer vision research

## When to choose
- you need photorealistic synthetic data with perfect ground truth labels
- you want to train or evaluate embodied AI agents in Unreal Engine worlds
- you need a Python API to interact with UE scenes without learning the engine
- you want ready-made virtual environments via UnrealZoo

## When to avoid
- you need lightweight simulation without a game engine dependency
- your project targets engines other than Unreal Engine
- you only need simple 2D image augmentation rather than 3D rendering

## Facets
- artifact type: plugin
- maturity: active
- function: simulation, computer-vision, image-processing, machine-learning, plugin-system, sdk
- domain: computer-vision, simulation, machine-learning, artificial-intelligence
- platform: windows, python, cpp, cross-platform
- tags: unreal-engine, synthetic-data, embodied-ai, ue4, ue5, virtual-worlds, ground-truth-annotation, mcp, game-development, linux, macos

## Member repositories
- unrealcv/unrealcv (main) score 67

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:26.212765+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:46:19.792821+00:00, confidence not recorded.
  - readme: https://github.com/unrealcv/unrealcv (fetched 2026-08-28T04:06:26.212765+00:00, sha af5222265e4d)
  - homepage: https://unrealcv.org (fetched 2026-08-29T10:26:38.057890+00:00, sha 1716b7469205)
  - site_page: http://docs.unrealcv.org (fetched 2026-08-29T10:26:38.066851+00:00, sha 9a313e69d10e)
  - site_page: http://docs.unrealcv.org/en/master/plugin/install.html (fetched 2026-08-29T10:26:38.068731+00:00, sha d9fed05923c4)
- Data as of 2026-08-30T08:39:29.467469+00:00.