Stable Diffusion AI Image Generator

For those who may not know, this extension for Auto1111 web-ui is extremely useful for managing models.


You can:
  • Cleaning/pruning models.
  • Converting to/from safetensors/ckpt.
  • Extracting/replacing model components (VAE, CLIP, UNET).
  • Identifying/debugging model architectures.
 
Yeah, long winded creative writing style descriptions doesn't do much in the way of SD prompts. If you rework the core idea of it into the standard SD type prompt then you get better results but you'll almost never get exactly what you want by just prompting alone.

Without spending too much time or using any other models \ add-ons on this, this was made just with prompting and stock SD1.5, it almost matches the description except for the silver and blue colour scheme and number of fingers.

View attachment 1521463

The prompt / negative prompt / steps / sampler, seed, etc data hereunder.

perfectly centered raw analog photo of a (robot:1.1) with a round head, (sleek:1.2), tall slender body, modern design, silver plating and blue accents, high tech, studio backdrop, photorealistic, 8k, studio lighting, award winning art, trending on artstation, concept art by Greg Rutkowski and Alphonse Mucha, highly detailed, cinematic lighting, highest quality, masterpiece, UHD, HDR, creative, Particularly Detailed,octane render, trending on Artstation,unreal engine dreamy, Outstandingly Detailed, lumen, raytracing, subsurface scattering, 8k, cinematic lighting, masterpiece,3D art by David Levy
Negative prompt: bulky, combat, chunky, fat, man, woman, child, human, portrait, closeup, bad art, boring, blurred, unclear, speech bubbles, text, captions, subtitles, titles, writing, credits, artifacts, compression, mediocre, average, uninspiring, bland, flat, mundane, run of the mill, overexposure, cropped, watermark, logo, barcode, UI, signature, text, label, error, title, stickers, markings, speech bubbles, cropped, lowres, low quality, artifacts, bad_prompt, bad-hands-5, EasyNegative
Steps: 25, Sampler: DPM++ 2M Karras, CFG scale: 10, Seed: 2891354958, Size: 512x768, Model hash: cc6cb27103, Model: model
Handpicked a few, of 6, from each to not shìt-up the thread.

Anything V4.5 - no woman? Get only big-titted women!
00006-2891354958.png00009-2891354961.png

DreamShaper V4 w/ Baked VAE
00023-2891354963.png00022-2891354962.png00020-2891354960.png00018-2891354958.png
 
Can it generate non humanoid things like houses or cars?
 
Damn, these AI image generators are fantastic.

I've been playing with Adobe Firefly but I really like the idea of an offline tool that I can feed existing images into.
 
Damn, these AI image generators are fantastic.

I've been playing with Adobe Firefly but I really like the idea of an offline tool that I can feed existing images into.
How long did it take to get access? I registered about a week ago.
 
There's a new ControlNet method available, called refererence only which is quite useful if you want to create AI variations from an original source picture. I haven't played with it much yet but it seems pretty cool.

 
There's a new ControlNet method available, called refererence only which is quite useful if you want to create AI variations from an original source picture. I haven't played with it much yet but it seems pretty cool.

Yeah this is pretty impressive.

Here's a photo I took :

Holiday01.jpg


And some AI variations of it using the new ControlNet Reference mode.

AI1.jpg

AI2.jpg
AI3.jpg
 
Combining the new Reference Controlnet model with a depth map from a photo, I created a new Bioshock Rapture-inspired digital painting. From what I've seen it may not even be necessary to train or use "style" models anymore as one can simply use reference images to capture a style, although one might want to use a secondary style such as 3D style, photo style or digital painting in which case those models still have their place in the workflow.

Reference Image :
Bioshock Ref.png

Depth Map from this pic:
City Pic.jpg

Result :
Rapture_S.jpg
 
In other news, the latest Beta version of Photoshop has some Firefly generative art capabilities now and it's fairly decent though it probably won't replace SD as a means of creating new images for a while yet, as far as I'm concerned. There's also some AI related enhancements to their touchup tools, selection tools and a remove tool that should save some time when working with cleanups and removing unwanted elements in photos.
 
Anyone tried using the "Tiled VAE" extension? It claims to essentially eliminate any out-of-memory issues and allows for extremely high resolution inference.
 
Anyone tried using the "Tiled VAE" extension? It claims to essentially eliminate any out-of-memory issues and allows for extremely high resolution inference.
I tried it... ran out of memory every.single.time...
 
Top
Sign up to the MyBroadband newsletter
X