@@WadeMcMaster I would love to learn a pipeline where I can get as much control over generating a tight, consistent collection of generated assets (heads, trees, etc.) - preferably in a customised style. I want to then build more complicated compositions that all have that 'look'. There's an artist out there called Von Woolfe who is doing this but his process is unknown to me. His work is outstanding.
I think the first step would be creating an image in the style you'd like, then use it as a style reference image - and make more from that. AS you create more, you have more images you can use to achieve something similar with also as a style reference. Keep the prompt of the style pretty consistent too.
Hey Wade. Is there a any AI that I can upload say a photo of my dog and then just use AI to regenerate him in any background or ask it to manipulate the image into a caricature?
With StylarAi, you can create a background, upload the dog image and remove the background, then with generative fill blend it in (like I did with the robot in the video). Or You could try Alpaca Chroma to edit the image around the dog: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-bFxsAyHey68.htmlsi=7PAnhLUc_K2QTU3q You can auto cut out the dog with Adobe express: www.adobe.com/express/feature/image/remove-background Then creat a background in chroma.and Import the dog over it. Then use generative mask to high the feet or edges that need blending in. You could also simply not use the generation mask in Chroma, and prompt for a caricature of a dog. I hope that helps!
I do get ya, but I believe it's two different AI algorithms - one for the initial image, one specifically for faces. Hope they are able to basically combine them soon enough
@@WadeMcMaster You should, you can do pretty much everything with comfyui that you can do here with custom nodes. Once I got passed the learning curb (which is no doubt a barrier) I have not seen anything else out there that I could justify paying for. In fact, if I'm honest all of these sites seem under powered in ability next to comfyui.