A field guide to the data-annotation jungle
Someone finally catalogued the 60+ tools for turning raw data into labeled training fodder, from CVAT to obscure point-cloud utilities.

What it does This is a curated awesome-list that sorts data-annotation tools by modality—image/video, text, multimodal/pointcloud—and license (open source vs. commercial). It also flags UI components you can embed in your own apps, plus a grab-bag of oddities like genomics annotators.
The interesting bit The maintainer puts skin in the game: they single out CVAT as the consensus pick for computer-vision tasks, based on independent agreement from “several other people in industry and academia.” That kind of editorial judgment is rare in awesome-lists, which usually just accumulate links.
Key highlights
- ~50 open-source tools and ~20 commercial ones, with modality tags (image, video, text, audio, 3D, pointcloud)
- Covers niche territory: DICOM medical imaging, synthetic-aperture radar, ROS bag files, genomics
- Includes embeddable components (React, Javascript) for rolling your own annotation UI
- Maintained actively enough that the README asks for PRs and issues
Caveats
- No comparison matrix, feature charts, or maturity ratings—you still have to click through each tool
- Some entries are just name + link + one-word tag; depth varies wildly
- The “best” recommendation (CVAT) is asserted, not benchmarked
Verdict Worth bookmarking if you’re shopping for annotation infrastructure or trying to avoid rebuilding a bounding-box editor for the hundredth time. Skip it if you already have a workflow you tolerate; this list won’t convince you to migrate.