#SwarmUI

Shanie MyrsTear

@shanie.tails.ch

8 months ago

A young mouse-humanoid woman in an anime style, standing on a sunny beach. She is holding a hamburger up in front of her with a surprised expression. The overall tone is bright, cheerful, and playful. She has long, black, wavy hair with turquoise tips. She has large, bright green eyes and a fair complexion. She is wearing a patriotic outfit consisting of a red and white striped top resembling the American flag, a white tiara with diamonds, blue denim shorts, and a white belt.

Mouse supports AMERICA by eating this burger (nice) and losing medicaid (oops)

#usa #us #aiart #comfyui #swarmui #4thofjuly #bigbeautifulbill

Dark Arts AI

@dark-arts-ai.bsky.social

9 months ago

Base Prompts Prompt: {value}, realistic, detailed, detailed background, depth of field Negative: 3d, anime, 2d , illustration, drawn, monochrome, abstract, low detail, worst quality, bad quality, low quality, lowres, wrong, bad, weird, wrong anatomy, bad anatomy, weird anatomy, wrong pov hands, bad pov hands, wrong hands, bad hands, extra hands, extra fingers, missing fingers, 6 fingers, 4 fingers, crossed eyes, lazy eye, heterochromia, text, username, scan, words, signs, censored, watermark, logo, artist_name, website_name, game_name, title, multiple views, error, jpeg artifacts, unfinished, displeasing, oldest, early

Base Prompts Prompt: {value}, realistic, detailed, detailed background, depth of field Negative: 3d, anime, 2d , illustration, drawn, monochrome, abstract, low detail, worst quality, bad quality, low quality, lowres, wrong, bad, weird, wrong anatomy, bad anatomy, weird anatomy, wrong pov hands, bad pov hands, wrong hands, bad hands, extra hands, extra fingers, missing fingers, 6 fingers, 4 fingers, crossed eyes, lazy eye, heterochromia, text, username, scan, words, signs, censored, watermark, logo, artist_name, website_name, game_name, title, multiple views, error, jpeg artifacts, unfinished, displeasing, oldest, early

My current SDXL setup:
SwarmUI

Base/Refiner/Detailer Model
Hyphoria (civitai.com/models/15958...)
CFG 4, Steps 25

Refiner: 0.15 Denoise, 2x Upscale (Step-Swap Method)

Detailers: 'face' and 'body'
0.4 Denoise

#sdxl #aiart #aiartist #aiartcommunity #swarmui #comfyui #promptart #synthart

Shanie MyrsTear

@shanie.tails.ch

10 months ago

Three characters walking in a snowy town scene, one poking between the others, looking at something.

Two characters facing away from the camera, taken from hip level, looking towards another character, in the distant center, walking up to them in greeting. center. IN the background, a festival of sorts it happening, confetti and balloons in the air.

One character, holding popcorn and looking at the other two characters, who are looking back at them. They're in an arcade.

I've been trying to integrate multiple characters into a scene without individually generating each or using ControlNet. I think they came out well so far, though backgrounds could be a bit better.

CW: Some booty focus in image 2. Aesthetic choice.

#ai […]

[Original post on mastodon.tails.ch]

Shanie MyrsTear

@shanie.tails.ch

1 year ago

Two girls with long hair, one adorned with a floral hair ornament. The one with the floral ornament, a redhead, looks back in slightl surprise to the other girl, silver haired, as the silver haired girl pokes her head over her shoulder to giver her a wink.

Was doing some more AI experimentation and, while doing some face manipulation, accidentally rendered just the faces in high quality. Results were pretty nice.

#aiart #aiartcommunity #aiartwork #anime #stablediffusion #swarmui

Shanie MyrsTear

@shanie.tails.ch

1 year ago

Quickly generated image, pretty low quality of an anime girl with blue hair giving a wink and peace sign wearing a black and purple bodysuit on top of a skyscraper.

Same as the previous image, except this time the body suit looks better.

Screenshot of the user interface of SwarmUI with the bodysuit highlighted in the Mask layer, and Init Image configured.

This weeks annoyance, beyond elections, was attempting to #inpaint using local #ai #image #generation system #swarmUI v.0.9.3.1.

Steps are as follows:
1) Generate an image
2) Select image, "Edit Image" button
3) Draw white on "mask" layer.
4) Click "Use […]

[Original post on mastodon.tails.ch]

Browncat AI

@browncat.bsky.social

1 year ago

#SwarmUI のRefinerで #SD3 Mediumを指定する手法で、 #Pony 系ベース（PonyRealism_v21VAE）の画像を生成してみました。
Refinerなしに比べると、顔の描写が改善される傾向にあるようです。