Megvii-BaseDetection / Megvii-BaseDetection/OTA
About the code
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 246
- Forks
- 24
- PR merge metrics
- No merged PRs in 30d
Description
What is the reason for this operation? "Rescale pi so that the max pi for each gt equals to 1."
_, pi = self.sinkhorn(mu, nu, loss)
378 # Rescale pi so that the max pi for each gt equals to 1. #一定需要rescale吗？
379 rescale_factor, _ = pi.max(dim=1)
383 pi = pi / rescale_factor.unsqueeze(1)
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
Start at the code around the self.sinkhorn(mu, nu, loss) call and inspect how pi is normalized after the transport assignment. Determine from the surrounding implementation and the OTA paper whether rescaling is required, then document the rationale or clarify the operation's intended behavior.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- computer-vision, machine-learning
- Issue type
- Documentation
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100