ComfyUI_pixelization

ComfyUI node that pixelizes images

Workflow preview

Original

Installation

1. Clone repository to ComfyUI/custom_nodes/ComfyUI_pixelization
2. Go to folder
3. Run python ./install.py
4. Download checkpoints to ComfyUI/custom_nodes/ComfyUI_pixelization/checkpoints
5. Use node Pixelization > Pixelization to generate pixelated image

Models

Download all three models from the table and place them into the checkpoints directory inside the extension at ComfyUI/custom_nodes/ComfyUI_pixelization/checkpoints.

url	filename
https://drive.google.com/file/d/1VRYKQOsNlE1w1LXje3yTRU5THN2MGdMM/view?usp=sharing	pixelart_vgg19.pth
https://drive.google.com/file/d/17f2rKnZOpnO9ATwRXgqLz5u5AZsyDvq_/view?usp=sharing	alias_net.pth
https://drive.google.com/file/d/1i_8xL3stbLWNF4kdQJ50ZhnRFhSDh3Az/view?usp=sharing	160_net_G_A.pth

Alternatively, you can download the models from the CivitAI mirror.

Credits

• AUTOMATIC1111 extension: https://github.com/AUTOMATIC1111/stable-diffusion-webui-pixelization
• Original repo: https://github.com/WuZongWei6/Pixelization
• Code I used for reference: https://github.com/arenatemp/pixelization_inference