[Suggestion] Acknowledging UCGM in both Code and Technical Report (as referenced in Saining Xie et al.)
Nobody has claimed this yet.
Assessment
- Difficulty
- 2/5
- Estimated time
- 1-3 hours
- Newbie friendliness
- 45/100
- Issue type
- Documentation
- Clarity
- Mostly clear
- Activity status
- Stale
- Tech stack
- python
- Domain
- documentation
Research direction
Start by reviewing UniPic-3/qwen_image_edit_fast/train_cm_dmd.py lines 79-88 and compare them with the linked UCGM implementation. Then inspect the README or Technical Report reference list; the work is done when the code attribution and UCGM reference requested in the issue are added.
Written by the indexing model from the issue text.
Description
Hello @SkyworkAI-Team,
Thanks for the great work on UniPic!
While going through the code, I noticed that the implementation of the Finite Difference technique in train_cm_dmd.py shares the same logic and specific variable naming as the UCGM repository.
UniPic: train_cm_dmd.py (L79-L88)
UCGM: unigen.py (L278-L291)
Since this technique is widely recognized as originating from UCGM, other recent works (e.g., Saining Xie et al. in arXiv:2601.09881) have explicitly cited it when using the same method.
To align with these community standards, I have two kind suggestions:
-
Add a code comment: It would be helpful to mark the source in the code (e.g., # Adapted from UCGM: https://github.com/LINs-lab/UCGM) so future developers know where this technique comes from.
-
Update References: Consider adding UCGM to the reference list in the README or Technical Report to acknowledge their contribution.
Hope this is helpful for refining the project!
- Dominant language
- Python
- Stars
- 875
- Forks
- 43
- PR merge metrics
- No merged PRs in 30d
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.
More from SkyworkAI/UniPic
-
Difficulty 4/5 3-5 days Newbie friendliness 25/100
-
Difficulty 5/5 Over a week Newbie friendliness 15/100
-
Difficulty 5/5 Over a week Newbie friendliness 25/100
-
Difficulty 5/5 Over a week Newbie friendliness 10/100
-
Inpainting Pipeline Open
Difficulty 5/5 Over a week Newbie friendliness 25/100
All issues in SkyworkAI/UniPic
Similar issues
-
Difficulty 2/5 1-3 hours Newbie friendliness 74/100
bancolombia/sentinel#23 ·
-
test md OpenCI
Difficulty 2/5 1-3 hours Newbie friendliness 74/100
-
integration:quickjs org:external priority:backlog topic:code-interpreter topic:middleware type:feature
Difficulty 2/5 1-3 hours Newbie friendliness 74/100
langchain-ai/deepagents#6450 ·
-
bug client
Difficulty 2/5 1-3 hours Newbie friendliness 88/100
-
Difficulty 2/5 1-3 hours Newbie friendliness 74/100