https://github.com/wasasquatch/portrait-generator-init-image-prep
Process portrait images to be more compatible with the Portrait Generator in Disco Diffusion or DD Portrait Playground for creating unique portraits.
https://github.com/wasasquatch/portrait-generator-init-image-prep
Last synced: about 1 year ago
JSON representation
Process portrait images to be more compatible with the Portrait Generator in Disco Diffusion or DD Portrait Playground for creating unique portraits.
- Host: GitHub
- URL: https://github.com/wasasquatch/portrait-generator-init-image-prep
- Owner: WASasquatch
- License: mit
- Created: 2022-08-01T01:17:49.000Z (about 4 years ago)
- Default Branch: main
- Last Pushed: 2022-08-19T06:38:45.000Z (almost 4 years ago)
- Last Synced: 2025-01-19T12:54:14.738Z (over 1 year ago)
- Language: Jupyter Notebook
- Size: 42 KB
- Stars: 0
- Watchers: 4
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Portrait Generator `init_image` Prep v1.0 

Process portrait images to be more compatible with the Portrait Generator in [Disco Diffusion](https://discodiffusion.com/) or [DD Portrait Playground](https://rebrand.ly/portrait-playground) for creating unique portraits.
## Why prep portrait `init_images`?
Prepping portrait images allows DD to start with an initial init that is easier to imrpovise on. When using a raw portrait of a person, DD has a hard time keeping the cohesion of the original image through diffusion.