wkentaro/labelme

Image annotation with Python. Supports polygon, rectangle, circle, line, point, and AI-assisted annotation.

73
/ 100
Verified

This tool helps data annotators efficiently label objects and regions within images or video frames for machine learning tasks. You input raw images or video, and it outputs detailed annotations saved as JSON files, outlining specific shapes like polygons, rectangles, circles, lines, or points. It's designed for anyone preparing visual datasets for training computer vision models.

15,641 stars. Actively maintained with 189 commits in the last 30 days.

Use this if you need to precisely mark objects, boundaries, or specific points in images or video to create training data for AI models.

Not ideal if you're looking for a fully automated annotation solution without any manual input or if your primary need is general image editing rather than data labeling.

data-labeling computer-vision image-annotation machine-learning-datasets AI-training
No Package No Dependents
Maintenance 22 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 25 / 25

How are scores calculated?

Stars

15,641

Forks

3,648

Language

Python

License

GPL-3.0

Last pushed

Mar 13, 2026

Commits (30d)

189

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/wkentaro/labelme"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.