[Fix] Set random seed for palette if not given.
#1152
Conversation
|
|
Codecov Report
@@ Coverage Diff @@
## master #1152 +/- ##
==========================================
- Coverage 90.11% 90.07% -0.04%
==========================================
Files 125 125
Lines 7262 7268 +6
Branches 1206 1206
==========================================
+ Hits 6544 6547 +3
- Misses 515 518 +3
Partials 203 203
Flags with carried forward coverage won't be shown. Click here to find out more.
Continue to review full report at Codecov.
|
| @@ -245,8 +245,12 @@ def show_result(self, | |||
| seg = result[0] | |||
| if palette is None: | |||
| if self.PALETTE is None: | |||
| # Fix colors by setting random seed. | |||
| state = np.random.get_state() | |||
There was a problem hiding this comment.
Can add more comments, refer to https://github.com/open-mmlab/mmdetection/blob/master/mmdet/core/visualization/image.py#L94-L96.
There was a problem hiding this comment.
Also, add this line in comments “The palette may be different in each iteration if not specified.”
There was a problem hiding this comment.
Also, add this line in comments “The palette may be different in each iteration if not specified.”
But if set np.random.seed(42), the random palette will be fixed in each iteration.
There was a problem hiding this comment.
Of course. My point is an additional straightforward explanation in comments for this modification.
|
Hi @jbwang1997 |
There was a problem hiding this comment.
https://github.com/open-mmlab/mmsegmentation/blob/master/mmseg/datasets/custom.py#L352
Here is another random function, you can modify it at the same time.
mmseg/models/segmentors/base.py
Outdated
| @@ -245,8 +245,16 @@ def show_result(self, | |||
| seg = result[0] | |||
| if palette is None: | |||
| if self.PALETTE is None: | |||
| # Get random state before set seed, | |||
| # and restore random state later. | |||
| # Prevent loss of randomness. | |||
There was a problem hiding this comment.
| # Prevent loss of randomness. | |
| # It will prevent loss of randomness, as the palette may be different in each iteration if not specified. |
There was a problem hiding this comment.
OK, I will check it quickly.
Junjun2016
changed the title
palette in iteration.palette if not given.
…lab#1152) * Fix colors * fix comments * Add comments * Add comments. Add random seed in datasets
Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily get feedback. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers.
Motivation
The palette may be different in each iteration if not specified.
Modification
Add np.random.seed(42) in base.py.
Following the method in mmdet.core.visualization.
BC-breaking (Optional)
Does the modification introduce changes that break the backward-compatibility of the downstream repos?
If so, please describe how it breaks the compatibility and how the downstream projects should modify their code to keep compatibility with this PR.
Use cases (Optional)
If this PR introduces a new feature, it is better to list some use cases here, and update the documentation.
Checklist