Generating pictures with the use of VQGAN & CLIP as an input text is given
"village landscape, made of lines sketch style river sun and trees"
"Transformation power of AI, supercharged DNA of the new digital age challenge"
Notebook generates images from text phrases with VQGAN and CLIP (Z + Quantize method with augmentations).
you dont need to give pictures to transfer styles from one to the other
Notebook initially was made by Katherine Crowson (https://github.com/crowsonkb, https://twitter.com/rivershavewings). The Biggan + CLIP method was made by https://twitter.com/Advadnoun. Aggregated explanations, and modifications by Eleiber # 8347, and the friendly interface was made thanks to Abulafia # 3734.
All tributes to above authors I just went through to see how quickly new GAN CLIP etc models are taking over sphere of art.Similarly like crating videas was previously reserved for big studios nowaday everybody with access to internet and handy could do it.
After a bit more than a year much more is possible Lets look at the project pf dreambooth
created by: lencz.sla@gmail.com