Thank you, that at looks much nicer than Logseq.
Unfortunately, I don’t have a good solution either. I am not even talking about tagging the images, I take screenshots during meetings, and it is already difficult to keep track of a handful of images, especially when they open and close when they get mouse focus. Sometimes screenshot comes first, sometimes annotation, so it is not very structured and hard to keep track.
The way I would like it is to have an image mode, where for blocks with a certain property images are shown on the left and corresponding text is on the side. Similar to the Powerpoint handout view:
This way one could quickly annotate and sort images.
Here is another relevant discussion: