COCO Annotator
7.4#5 in ai image annotation toolsFree planFree planWebAPILinuxSelf-hosted
Visit| Fact | COCO Annotator | MakeSense.ai |
|---|---|---|
| Annotation features | It supports object segmentation, keypoints, disconnected objects as one instance, multiple labels per image segment, and custom metadata. | The tool supports point, line, rectangle, polygon, and label annotations. |
| Praise: usability | The ACM review describes quick navigation, bounding-box, polygon, and eraser tools, and straightforward export. | A 2025 Journal of Applied Crystallography paper describes the interface as intuitive and says its one-by-one manual labeling speed was comparable to the authors’ pipeline. |
COCO Annotator suits teams preparing labeled image data for object detection or localization. It is for users comfortable running a self-hosted service with Docker and docker-compose.
MakeSense.ai suits people preparing labeled datasets for small computer vision projects. It may also fit users who want browser-based annotation and local or Docker setup options.