-
Notifications
You must be signed in to change notification settings - Fork 9.4k
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
More robust Accuracy module #3155
Conversation
Johnson-Wang
commented
Jun 30, 2020
- Make mmdet.models.Accuracy class more robust to empty predictions.
- Add score threshold for Accuracy calculation
- Add unittest for Accuracy module
- change accuracy to pos_recall in FSAF and use Accuracy module to calculate it.
…to pos_recall # Conflicts: # mmdet/models/dense_heads/fsaf_head.py
mmdet/models/losses/accuracy.py
Outdated
topk (int | tuple[int], optional): If the predictions in ``topk`` | ||
matches the target, the predictions will be regarded as | ||
correct ones. Defaults to 1. | ||
thresh (float, optional): If not None, threshold for pred with scores | ||
under this value being incorrect. Default to None. |
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.
This sentence is a little confusing. Is it better to say this?
If not None, predictions with scores under this threshold are to be considered incorrect
mmdet/models/losses/accuracy.py
Outdated
"""Module to calculate the accuracy | ||
|
||
Args: | ||
topk (tuple, optional): The criterion used to calculate the | ||
accuracy. Defaults to (1,). | ||
thresh (float, optional): If not None, threshold for prediction | ||
scores under this value being incorrect. Default to None. |
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.
Need to also modify here.
* change accuracy to positive recall in FSAF * use mmdet accuracy module to calculate pos recall in fsaf * Change label to long in unittest * revise docstring according to comments * revise docstring according to comments