In deep learning based computer vision tasks, are there any models/tools that we can use to predict bounding box coordinates of a region given that an object/entity is present i