Target annotation is a core data annotation method for object detection and recognition tasks, which locates and marks specific target objects (such as "car", "pedestrian", "traffic light", "text area") in data (mainly images and videos), clarifying the position, category or attribute information of the target.
