Fake-QRcode ControlNet

These are ControlNet checkpoints trained on runwayml/stable-diffusion-v1-5 to generate recognizable AIGC QRcode image.

Model Details

Details will be added soon...

Use with diffusers

See the snippet below for usage with diffusers:

import cv2
import numpy as np
import torch
import os, sys

from diffusers import StableDiffusionControlNetPipeline, ControlNetModel, AutoencoderKL, EulerAncestralDiscreteScheduler
from PIL import Image

controlnet = ControlNetModel.from_pretrained("ghoskno/Fake-Qrcode")
pipe = StableDiffusionControlNetPipeline.from_pretrained(
  "runwayml/stable-diffusion-v1-5", controlnet=controlnet, torch_dtype=torch.float16
)
pipe.scheduler = DDIMScheduler.from_config(pipe.scheduler.config)
pipe.enable_model_cpu_offload()

generator = torch.manual_seed(412052000)

qrcode = cv2.imread('path_to_qrcode.png')
qrcode = cv2.resize(255 - qrcode, (1024, 1024))

image = pipe(
  "Blooming chinese chrysanthemum, green leaves growing wantonly, flowers, Complex patterns on the border, Masterpiece Art, Beauty, 8K,  Unreal Engine",
  Image.fromarray(qrcode),
  generator=generator,
  num_inference_steps=37,
  guidance_scale=7,
  controlnet_conditioning_scale=1.85
).images[0]

Some examples

input qrcode image

prompt: Blooming chinese chrysanthemum, green leaves growing wantonly, flowers, Complex patterns on the border, Masterpiece Art, Beauty, 8K, Unreal Engine

prompt: Plum blossoms in the snow, pink stamens, green leaves and branches growing wantonly, flowers, Complex patterns on the border, Masterpiece Art, Beauty, 8K, Unreal Engine

Limitations and Bias

• No strict control by input prompt
• Sometimes generate confusion or generate unrecognizable QRcode images