-
Notifications
You must be signed in to change notification settings - Fork 1.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Feature] Support random degradations during training #504
Conversation
Codecov Report
@@ Coverage Diff @@
## master #504 +/- ##
==========================================
+ Coverage 80.41% 80.54% +0.13%
==========================================
Files 191 193 +2
Lines 10378 10759 +381
Branches 1553 1631 +78
==========================================
+ Hits 8345 8666 +321
- Misses 1796 1835 +39
- Partials 237 258 +21
Flags with carried forward coverage won't be shown. Click here to find out more.
Continue to review full report at Codecov.
|
A general note: for augmentations, please add the added/modified keys in docstring. |
self.keys = keys | ||
self.params = params | ||
|
||
def _apply_gaussian_noise(self, input_): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
How about input_
-> input
?
* AddRandomDegradations * Add blur_kernels.py * Refactor random degradations * Minor fix and add unittest * update, pull from master * Update * Add license * minor fix
Motivation
Recent works on face restoration or real-world super-resolution rely heavily on applying random degradations during training.
Modification
This PR adds support to generate random degradations in the data processing pipeline. It currently supports
This is a generic class. Supporting other degradations can be done by adding corresponding operations in this class.