invoke-ai / invoke-ai/InvokeAI
[enhancement]: Add "Bicubic" option to Upscaler
- Dominant language
- Python
- Stars
- 28.2k
- Forks
- 3k
- Avg merge
- 6d 5h
- Merged PRs (30d)
- 19
Description
### Is there an existing issue for this?
- [x] I have searched the existing issues
### Contact Details
_No response_
### What should this feature add?
When using the Upscaler, selecting the right Upscale Model can be tricky since they all have issues of their own, such as artifacting, discoloration, etc. Often times, that part of the upscaling process is less important than the prompt and SDXL model settings.
Bicubic upscaling is a traditional (pre-AI) way of upscaling while simply smoothing the pixel aliasing, without adding artifacts or changing colors or other aspects of the image. Such an option could make upscaling certain kinds of images much less challenging.
### Alternatives
_No response_
### Additional Content
_No response_
Contributor guide
No contributing guide indexed for this repository
Research direction
No files, tests, or entry points are named. Start by locating the Upscaler selection and processing path, then determine where a non-AI Bicubic choice belongs. Done means Bicubic is selectable for upscaling and produces the intended smoothing without introducing the artifacts or color changes described in the issue.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100