Annotation completeness means the degree to which annotation results cover all information to be annotated in data (e.g., all targets are annotated in object detection), which is one of the core indicators of annotation quality and directly affects the model's comprehensive learning of data.
