Support ScaledLoss #141
Merged
Support ScaledLoss #141
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
efernandez
commented
Feb 13, 2020
This was referenced Feb 13, 2020
Merged
Merged
ayrton04
approved these changes
Feb 20, 2020
efernandez
force-pushed
the
scaled_loss
branch
from
6e4a520
to
f446ede
Compare
svwilliams
approved these changes
Feb 20, 2020
efernandez
force-pushed
the
scaled_loss
branch
from
f446ede
to
1ddcf0a
Compare
|
@svwilliams I just addressed all the typos and the test suggestions. Thanks. |
svwilliams
approved these changes
Feb 21, 2020
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Outdated suggestions cannot be applied.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.
This adds supports to
ceres::ScaledLoss: https://github.com/ceres-solver/ceres-solver/blob/032d5844c2db412e25f780c6acf95ecc7a0b7975/include/ceres/loss_function.h#L311-L349This allows to scale the output of a loss function by a scalar factor
a.This is useful if we want to scale the loss function influence of a particular source more than others, or even to workaround the incorrect implementation of the
TukeyLoss: https://github.com/ceres-solver/ceres-solver/blob/032d5844c2db412e25f780c6acf95ecc7a0b7975/internal/ceres/loss_function.cc#L118-L132, that should be multiplied by2because the cost multiplies the output of the loss functionEvaluatemethod by0.5, which is Ceres solver convention.