Computer vision dataset Open Images V7 adds point labels, which can be used for instance segmentation model training

Google further expands the computer vision dataset Open Images, releasing the latest V7 version, adding a new annotation type called Point Labels. In addition, Open Images V7 also adds an all-in-one visual tool for users to use multiple tools at the same time. Annotate the data to better explore the data set.