【SD3辅助工具推荐】InstantX发布了三种SD3专属的ControlNet模式——Pose、Canny和Tile

InstantX 是一家专注于人工智能内容生成的独立研究机构。此前，曾开源著名的InstantID和论文《InstantID : Zero-shot Identity-Preserving Generation in Seconds》。随着本月12号，Stability AI正式开源了其产品 Stable Diffusion 3，这家机构又为SD的死忠粉带来了三款SD3专属的ControlNet工具——SD3-Controlnet-Pose、SD3-Controlnet-Canny 和 SD3-Controlnet-Tile。

SD3-Controlnet-Pose

control image	weight=0.0	weight=0.3	weight=0.5	weight=0.7	weight=0.9

请确认版本diffusers >= 0.30.0.dev0

Demo

import torch
from diffusers import StableDiffusion3ControlNetPipeline
from diffusers.models import SD3ControlNetModel, SD3MultiControlNetModel
from diffusers.utils import load_image

# load pipeline
controlnet = SD3ControlNetModel.from_pretrained("InstantX/SD3-Controlnet-Pose")
pipe = StableDiffusion3ControlNetPipeline.from_pretrained(
    "stabilityai/stable-diffusion-3-medium-diffusers",
    controlnet=controlnet
)
pipe.to("cuda", torch.float16)

# config
control_image = load_image("https://huggingface.co/InstantX/SD3-Controlnet-Pose/resolve/main/pose.jpg")
prompt = 'Anime style illustration of a girl wearing a suit. A moon in sky. In the background we see a big rain approaching. text "InstantX" on image'
n_prompt = 'NSFW, nude, naked, porn, ugly'
image = pipe(
    prompt, 
    negative_prompt=n_prompt, 
    control_image=control_image, 
    controlnet_conditioning_scale=0.5,
).images[0]
image.save('image.jpg')

SD3-Controlnet-Canny

control image	weight=0.0	weight=0.3	weight=0.5	weight=0.7	weight=0.9

请确认版本diffusers >= 0.30.0.dev0

Demo

import torch
from diffusers import StableDiffusion3ControlNetPipeline
from diffusers.models import SD3ControlNetModel, SD3MultiControlNetModel
from diffusers.utils import load_image

# load pipeline
controlnet = SD3ControlNetModel.from_pretrained("InstantX/SD3-Controlnet-Canny")
pipe = StableDiffusion3ControlNetPipeline.from_pretrained(
    "stabilityai/stable-diffusion-3-medium-diffusers",
    controlnet=controlnet
)
pipe.to("cuda", torch.float16)

# config
control_image = load_image("https://huggingface.co/InstantX/SD3-Controlnet-Canny/resolve/main/canny.jpg")
prompt = 'Anime style illustration of a girl wearing a suit. A moon in sky. In the background we see a big rain approaching. text "InstantX" on image'
n_prompt = 'NSFW, nude, naked, porn, ugly'
image = pipe(
    prompt, 
    negative_prompt=n_prompt, 
    control_image=control_image, 
    controlnet_conditioning_scale=0.5,
).images[0]
image.save('image.jpg')

SD3-Controlnet-Tile

control image	weight=0.0	weight=0.3	weight=0.5	weight=0.7	weight=0.9

请确认版本diffusers >= 0.30.0.dev0

Demo

import torch
from diffusers import StableDiffusion3ControlNetPipeline
from diffusers.models import SD3ControlNetModel, SD3MultiControlNetModel
from diffusers.utils import load_image

# load pipeline
controlnet = SD3ControlNetModel.from_pretrained("InstantX/SD3-Controlnet-Tile")
pipe = StableDiffusion3ControlNetPipeline.from_pretrained(
    "stabilityai/stable-diffusion-3-medium-diffusers",
    controlnet=controlnet
)
pipe.to("cuda", torch.float16)

# config
control_image = load_image("https://huggingface.co/InstantX/SD3-Controlnet-Tile/resolve/main/tile.jpg")
prompt = 'Anime style illustration of a girl wearing a suit. A moon in sky. In the background we see a big rain approaching. text "InstantX" on image'
n_prompt = 'NSFW, nude, naked, porn, ugly'
image = pipe(
    prompt, 
    negative_prompt=n_prompt, 
    control_image=control_image, 
    controlnet_conditioning_scale=0.5,
).images[0]
image.save('image.jpg')

感谢大家花时间阅读我的文章，你们的支持是我不断前进的动力。期待未来能为大家带来更多有价值的内容，请多多关注我的动态！