LabelU

What LabelU is

LabelU is an open-source platform for annotating images, video, and audio. Image work can include 2D bounding boxes, semantic segmentation, polylines, and keypoints; video and audio tools support segmentation, classification, and information extraction. Users can load pre-annotated data and adjust it. For images, AI services can detect and segment objects, run batch annotation, and report progress in real time. Reference servers are provided for Florence-2, GroundingDINO with SAM ViT-B, and SAM 3. Files can be imported from S3-compatible storage such as AWS S3 and MinIO, and listed export formats include JSON, COCO, and MASK. LabelU is available through an online instance and browser toolkit, or can be installed locally with pip and run at localhost:8000. Local setup specifies Python 3.11, SQLite by default, and optional MySQL. The project is published under the Apache 2.0 license by OpenDataLab. Reference model requirements include about 4 GB of memory for Florence-2 and GroundingDINO with SAM ViT-B, and about 8 GB for SAM 3, which also requires CUDA 12.6 or later.