In principle, I understand the reasons for the rejection of some of my work. But in this case, not so much! Could there be an explanation for this? Maybe the moderator made a mistake? )
I apologize if I wrote the wrong message in the wrong place.
Generally speaking, the images are too small to make an in-depth analysis of the images. But even at this size, you see multiple issues. For example:
I am not convinced that the cars per se have an IP problem. Except when the radiator figure is a bit more detailed in one of the pictures. The quality issues are, however, correct.
ABAMBO | Hard- and Software Engineer | Photographer
AI is trained on actual images. Where those images are sourced is anyone's guess. So probably not as unique as you think. 😉
Start with high resolution diffusions. It costs more to generate high-res images but Stock won't accept anything less.
AI is imperfect especially with fine details. It takes artistic talent to find & fix all AI imperfections. Some people can do it, some can't. Discard any you can't fix.
Generative AI Illustrations could take 1-3 months to be reviewed. Make the most of every AI image you submit. Focus on quality instead of quantity.
Generally speaking, the images are too small to make an in-depth analysis of the images. But even at this size, you see multiple issues. For example:
I am not convinced that the cars per se have an IP problem. Except when the radiator figure is a bit more detailed in one of the pictures. The quality issues are, however, correct.
Translate
ABAMBO | Hard- and Software Engineer | Photographer
From my perspective, these have both IP issues and Quality issues.
The cars look pretty detailed and distinct to me. I'm assuming this to be an IP issue.
These are some of the better A.I. illustrations I've seen, but there are noticeable errors if you study the details. The changes between textures are noticeable in all 3 shots in my opinion. Some of the silhouettes have artefacts around them.
Again, some of the better ones that I've seen. Cheers!
Thank you very much! Yes, of course there are mistakes. It seemed to me that this is normal for generating AI. I have seen accepted images with much bigger problems. For example, pictures on the topic "Boeing AI" 🙂
Yes, there are a lot of flawed AI images in the database, but I think Moderators are getting more skilled at detecting flaws and will reject images for flaws that might have gotten accepted months ago.