mmcv
mmcv copied to clipboard
[Feature] Add TTA transform
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
Transforms used by test time augmentation.
Modification
Please briefly describe what modification is made in this PR.
BC-breaking (Optional)
Does the modification introduce changes that break the back-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)
transforms = [
[dict(type='Resize', scale=(1333, 800), keep_ratio=True),
dict(type='Resize', scale=(1333, 800), keep_ratio=True)],
[dict(type='RandomFlip', prob=1.),
dict(type='RandomFlip', prob=0.)],
[dict(type='Normalize', mean=(0, 0, 0), std=(1, 1, 1))]]
tta_transform = TestTimeAug(transforms)
Checklist
- Pre-commit or other linting tools are used to fix the potential lint issues.
- The modification is covered by complete unit tests. If not, please add more unit test to ensure the correctness.
- If the modification has potential influence on downstream projects, this PR should be tested with downstream projects, like MMDet or MMCls.
- The documentation has been modified accordingly, like docstring or example tutorials.
Maybe use a cartesian product to implement the subroutines
from itertools import product
@TRANSFORMS.register_module()
class TestTimeAug(BaseTransform):
def __init__(self, transforms: list):
self.subroutines = [Compose(subroutine) for subroutine in product(*transfroms)]
def transform(self, results: dict) -> dict:
results_list = [] # type: ignore
for subroutine in self.subroutines:
routine_results = subroutine(deepcopy(results))
results_list.append(routine_results)
aug_data_dict = {
key: [item[key] for item in results]
for key in results_list[0]
}
return aug_data_dict