DAVIS: Densely Annotated VIdeo Segmentation

• Vision Dataset