Softologyさんのイラストまとめ

"A Winter Landscape" collection is now available to purchase on OpenSea. 10 unique and original images available.
https://t.co/yXCBd7ps7U
Thanks for looking and please feel free to retweet.

Multi-Perceptor VQGAN+CLIP v3 Text-to-Image script from @remi_durant is now included with Visions of Chaos.
https://t.co/5DpPdJnCD8
"a clown made of clay"
"a peacock made of metal"
"lineart of Big Bird by Alesso Baldovinetti"
"vector art of a forest fire"

Multi-Perceptor VQGAN+CLIP v2 Text-to-Image script from @remi_durant is now included with Visions of Chaos.
https://t.co/5DpPdJnCD8
"a babbling brook by Zhou Wenjing"
"a bedroom by Francesco Furini"
"a storybook illustration of a kitchen"
"vector art of Darth Vader"

Multi-Perceptor CLIP Guided Diffusion Secondary Model Method from @Somnai_dreams
now included with Visions of Chaos.
https://t.co/5DpPdJnCD8
"a 3D render of Robocop"
"an evil clown"
"Indiana Jones"
"zombies"

CLIP Guided Diffusion Secondary Model Method from @RiversHaveWings now included with Visions of Chaos
https://t.co/LbpHZsYlCc
"a nightmare creature"
"a photorealistic painting of The Grinch"
"a portrait of a woman"
"an art deco painting of a sad clown"

Hypertron Text-to-Image now included in Visions of Chaos.
https://t.co/LbpHZsYlCc
"a matte painting of a forest clearing"
"a renaissance painting of a ghost"
"a sea monster made of metal"
"a watercolor painting of a dragon"

AnimeGANv2 now included with Visions of Chaos.
https://t.co/oacfxXsQml
Here are example outputs processing Scarlett Johansson with each of the four AnimeGANv2 models.
AnimeGANv2 is very fast (seconds) and can handle large images without needing a GPU with loads of VRAM.

PyramidVisions Text-to-Image added to Visions of Chaos
https://t.co/LbpHZsYlCc
"a lush rainforest"
"a nightmare creature"
"an acrylic painting of an alien forest"
"Medusa"

DirectVisions Text-to-Image added to Visions of Chaos
https://t.co/3nUcgIS6of
"a pencil sketch of Buzz Lightyear"
"a rough seascape"
"an evening landscape"
"a stock photo of a president"

@pbaylies Just to be clear, these latest more realistic portrait results come from this notebook by Dango233 https://t.co/BILxsykfOR
which is a newer version of the original CLIP Guided Diffusion and Quick CLIP Guided Diffusion notebooks.
Anyone interested needs to use Dango233's script.