fix potential wrong reg_target assignment #52
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Fix potential wrong reg_target assignment on the center point collision problem as Objects as Points 6.1.1 said. Though we cannot forbid the center point collision problem, we can guarantee that the reg_target assignment behavior is deterministic, assigning to the nearest one.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fix potential wrong reg_target assignment on the center point collision problem as Objects as Points 6.1.1 said.
Though we cannot forbid the center point collision problem, we can guarantee that the reg_target assignment behavior is deterministic, assigning to the nearest one.
I'm not sure about the effect of this change, I have no enough machines to test it.
But I think this change is reasonable, especially for large epochs training.
By the way, I also wonder it is reasonable or not to remove this line
dist2[is_peak] = 0and believe the probability totally.Wonderful work! CenterNet2 tells us the power of probabilistic on two-stage detection.
But the CenterNet* could be further optimized, such as ground truth auto-assignment on feature pyramid instead of configuring scale range manually.