supervision
by roboflow
Supervision is a Python library of reusable computer vision tools. It helps you draw detection results on images and video, count objects in zones, and load, split, merge and convert image datasets between formats such as YOLO, COCO and Pascal VOC.
- Stars
- 51.4k
- Licence
- MIT
- Runs on
- Your computer
- Setup
- Moderate
- Language
- Python
- Age
- 1432 days
What you can do with it
- Draw boxes, masks and labels from any detection model onto images and video frames
- Count and track objects in defined zones, for example measuring customer dwell time in a shop
- Load, split, merge and convert image datasets between YOLO, COCO and Pascal VOC formats
Make money with it
A freelancer could offer custom video analytics, such as counting people or vehicles, to local shops, car parks or warehouses.
An idea, not financial advice. Check the licence before charging for anything built on it.
Ways a business could use it
- Build in-house footfall, queue or dwell-time analysis from existing camera footage without buying a specialist analytics product
- Prepare and convert labelled image datasets for custom detection models, such as stock or product checks
Can you use it commercially?
Free to use, change and sell, including in paid products. Keep the copyright notice.
Not legal advice. Read the project's licence file before building a product on it.
Getting it running
Install it with "pip install supervision" in a Python 3.10 or newer environment, then write Python code that uses it alongside a detection model.
Follow the official instructions in the project README.
Things to check
- Needs programming knowledge and a separate detection model; it is a library, not a finished app
- Some examples use the Roboflow Inference connector, which needs a Roboflow API key
- A GPU is usually helpful for running the models in real time, though the library itself does not require one
Who's talking about it
- datnguyenquy94 (news-radar): AI Open Source Trends 2026-10-30 (Issue #926) Featured in a roundup
Checked 30 Oct 2026. Summary drafted by AI from the project's README.
