←back to thread

67 points arduinomancer | 3 comments | | HN request time: 0.22s | source

I’ve noticed a lot of the time you can tell an image is AI generated because it has a shiny/glossy lighting look to it.

Has anyone figured out why this is the case?

1. LeoPanthera ◴[] No.41263249[source]
AI is compression and compression, of any lossy kind, usually works by removing the high-frequency information first. That applies to both audio and imagery. It's obviously not the only factor, but I bet it's an important one.
replies(2): >>41263297 #>>41263384 #
2. aitchnyu ◴[] No.41263384[source]
Is there any technique to reapply high frequency info to make stuff look realistic?
replies(1): >>41264942 #
3. Jensson ◴[] No.41264942[source]
With less lossy compression, sure. Meaning bigger models.