CVAT

CVAT (Computer Vision Annotation Tool), originally built by Intel, is an open-source, web-based platform for labeling images and video used to train computer vision models — bounding boxes, polygons, keypoints, segmentation masks — with support for multi-annotator workflows and quality control. It can be self-hosted for data security, which matters a lot to teams working with sensitive imagery they can't send to a third-party labeling service.