Seconding Bookstack. I’ve embedded videos in it and I don’t recall anything special to do it. I also think there’s a way to comment on specific pages…mostly because I remember disabling that functionality.
Agreed on the roles and permissions aspect though. It’s pretty standard to do that for bigger deployments, but it may be a bit overkill for a single user instance.
I run calibre off my desktop. You can enable the Calibre content server and it can serve up your books for download (or provide a web reader).
If you have an Android device, you can use something like Moon Reader (or any other reading app that supports epub or Pdf) to download content from the Calibre content server.
With respect to covers and metadata, Calibre can tag and fill in this info as well - out of the box it will scrape information from Amazon.
This is a great use of tech. With that said I find that the lines are blurred between “AI” and Machine Learning.
Real Question: Other than the specific tuning of the recognition model, how is this really different from something like Facebook automatically tagging images of you and your friends? Instead of saying "Here’s a picture of Billy (maybe) " it’s saying, “Here’s a picture of some precancerous masses (maybe)”.
That tech has been around for a while (at least 15 years). I remember Picasa doing something similar as a desktop program on Windows.