T-Rex Label is an out-of-the-box online data annotation tool. Featuring efficient and user-friendly AI-assisted annotation, coupled with an intuitive user interface, it supports both manual and automatic annotation, dedicated to helping users quickly build datasets for complex scenarios.
Its core advantages include:
(1) Out-of-the-box functionality with no need for download or installation—operated directly in the browser, significantly lowering the barrier to use;
(2) Automatic annotation powered by the state-of-the-art DINO-X vision model, balancing precision and efficiency;
(3) Exclusive T-Rex2 model with visual prompt support, excelling in rare object recognition and dense scene object detection, addressing pain points of traditional tools.
Currently, T-Rex Label mainly focuses on bbox and mask annotation, supports annotation of image and video data, and is highly compatible with mainstream COCO and YOLO data formats.