[Feature] Add test time augmentation base model.

[HAOCHENYE]

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

Add base model for test time augmentation.

Modification

If we want to use TTA model based on MMEngine, we should implement a model inherited from BaseTTAModel like this:

class CustomTTAModel(BaseTTAModel):
    def merge_preds(self, data_samples_list):
        """ merge data_sample_list to one. """
        ...

and then, TestTimeAug transform in mmcv should be used in test_pipeline:

file_client_args = dict(backend='disk')
tta_pipeline = [
        dict(type='LoadImageFromFile', file_client_args=file_client_args),
        dict(type='TestTimeAug',
             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='LoadAnnotations', with_bbox=True)],
                 [dict(type='PackDetInputs',
                       meta_keys=('img_id', 'img_path', 'ori_shape',
                                  'img_shape', 'scale_factor', 'flip',
                                  'flip_direction'))]])
    ]

TestTimeAug and CustomTTAModel are a combo, and should be used together.

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

1. Pre-commit or other linting tools are used to fix the potential lint issues.
2. The modification is covered by complete unit tests. If not, please add more unit test to ensure the correctness.
3. If the modification has potential influence on downstream projects, this PR should be tested with downstream projects, like MMDet or MMCls.
4. The documentation has been modified accordingly, like docstring or example tutorials.

[codecov]

Codecov Report

Base: 77.61% // Head: 77.67% // Increases project coverage by +0.05% 🎉

Coverage data is based on head (247a656) compared to base (5d7527d).
Patch coverage: 76.74% of modified lines in pull request are covered.

❗ Current head 247a656 differs from pull request most recent head a738cce. Consider uploading reports for the commit a738cce to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #538      +/-   ##
==========================================
+ Coverage   77.61%   77.67%   +0.05%     
==========================================
  Files         116      117       +1     
  Lines        8641     8677      +36     
  Branches     1785     1793       +8     
==========================================
+ Hits         6707     6740      +33     
- Misses       1632     1633       +1     
- Partials      302      304       +2     

Flag	Coverage Δ
unittests	77.67% <76.74%> (+0.05%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
mmengine/model/wrappers/distributed.py	51.51% <0.00%> (+2.94%)	⬆️
mmengine/runner/checkpoint.py	44.00% <50.00%> (ø)
mmengine/model/wrappers/test_time_aug.py	81.08% <81.08%> (ø)
mmengine/model/__init__.py	75.00% <100.00%> (ø)
mmengine/model/wrappers/__init__.py	70.00% <100.00%> (+3.33%)	⬆️
mmengine/model/base_model/base_model.py	86.11% <0.00%> (+2.77%)	⬆️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.