Hello. I am interested in applying some of the pre-trained deep learning packages to very high resolution (5-8cm) imagery I have, particularly for land classification. However, it seems like most of the available models were trained on Sentinel or Landsat. There is one "High Resolution land cover model" trained on 80cm to 100cm imagery in the Chesapeake Bay Region. Before I start diving into trying anything, I wanted to ask if this is an appropriate use of these models given my higher resolution imagery, or if I need to start from scratch? In general, do models need to be trained specifically for the resolution they will be looking at, or can they handle anything above the original resolution they were trained on? It would be great if these models work on very high resolutions as imagery is greatly improving in this area (i.e. drones). Tagging @BernSzukalski as he seems to be a moderator on this space. Thanks.