@vgan @maltimore I agree. I think you can see that when you prompt these networks to output faces. I couldn't get vqgan+clip to output a face that didn't look very distorted but the latent diffusion model im currently using gives decent results at least for some celebrity name in the prompt
#vqgan #clip #LatentDiffusion #text2art

I've been playing with text2image networks again after 4 months or so. I was quite happy with the results #LatentDiffusion #text2art #text2image @vgan @maltimore

RT @altsoph: Okay, let's try something new! (again, these experiments are based on CLIPIT/PixelDraw by @dribnet + a bunch of my demoscene hacks) #pixelart #generative #aiart #generativeart #digitalart #text2art [escher's tiling] + color boost + smoothness + spiral-zoom magic 2021-09-14 15:15:50 UTC

RT @altsoph: [The Starry Night in style of Van Gogh] + color boost + smoothness + parallax magic + endless loop #pixelart #generative #aiart #generativeart #digitalart #text2art 2021-09-06 15:47:33 UTC

RT @altsoph: [misty mountains landscape] + color boost + smoothness + parallax magic + endless loop Originally based on #CLIP by @OpenAI and PixelDraw by @dribnet. #pixelart #generative #aiart #generativeart #digitalart #text2art 2021-09-06 15:30:34 UTC