lllyasviel / lllyasviel/ControlNet
[FEATURE] I need MORE CONTROL over outpainting. Can you make a ControlNet level of control for OUTpainting (using an input image)
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 34.1k
- Forks
- 3k
- PR merge metrics
- No merged PRs in 30d
Description
Hello,
Is it possible to use an input image as a reference for controling how the outpaining image will be?
Let me give an example, let's take this image:
I would like to outpaint the right side but I want to use an image to control what will be outpainted:
So the input image I have to use is this one:
Same thing with this example:
Let's suppose I did not have any idea what is the rest of the image, then we can input an image to control what is the outpaining gonna do:
Outpainted image:
The actual outpaint method you described with the use of the llama thing is random and has not control of what is going to be outpainted except for the prompt and usual parameters.
Can you make something like what I described?
Thanks
@lllyasviel
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
No files, tests, or entry points are named. Start by locating the existing outpainting implementation and reviewing how its input image, mask, prompt, and control parameters are handled. Done would mean defining and implementing reference-image control for outpainting, with tests or examples showing that the requested region follows the supplied image.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100