Hello everyone, hope you are all doing fine. Thank you so much for the wonderful support and encouragement:).
I have been spending the last few days trying to generate portraits and modify them with AI. Like I have said, ArtBreeder is the best to generate portraits but my mind is slowly veering towards using StyleGAN3+CLIP as a tool to generate faces, not just for applying effects on ArtBreeder portrait results.
The image you see above was generated in StyleGAN3+CLIP using just text prompts and not a target image. Here's the video :
Another similar video although I gave an entirely different prompt :
Also, ArtBreeder Chaos gene/setting can be used to good effect to generate random changes which may sometimes fit our requirements if tweaking every gene is taking too much time.
Here are some results :
Video when used in StyleGAN3+CLIP :
The images can be made to look as realistic or artistic as you want :
And I did continue experimenting some more with DeepStyle on DeepDreamGenerator :
We can, of course, use the style of StyleGAN images as Deep Style style to generate images like this :
StyleGAN is a hit and miss sometimes when using target images instead of text prompts and can produce a wonderful soup of colours like this, though that was not what I was looking for :
That's all for today, hope this post was colourful and pleasing to the eyes and maybe gave a little information on producing portraits with AI😉 . Thank you for reading, see you in the next one!