I’m thinking about adding new features. Which one is more useful and should come next: searching with an image of an object, detecting changes over time using Sentinel-2 data, or detecting object in all Google Street View images within a selected area?