[Fix] flip_ratio in RandomAffine pipeline is inverted

[wdmwhh]

original: flip_ratio=1. means always no flip and flip_ratio=0. means always flip
after: flip_ratio=1. means always flip and flip_ratio=0. means always no flip

[wangruohui]

Because flip is used as a multiplier in line 479 and 480, so -1 stands for flip and 1 stands for no flip. Thus flip should be an 'inverse' flag as the result of the comparison.

Is it correct?

[codecov]

Codecov Report

Merging #799 (cd1328d) into master (696e3a8) will increase coverage by 0.02%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #799      +/-   ##
==========================================
+ Coverage   83.05%   83.08%   +0.02%     
==========================================
  Files         216      216              
  Lines       12192    12253      +61     
  Branches     1964     1978      +14     
==========================================
+ Hits        10126    10180      +54     
- Misses       1763     1764       +1     
- Partials      303      309       +6     

Flag	Coverage Δ
unittests	83.04% <100.00%> (+0.03%)	⬆️

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

Impacted Files	Coverage Δ
mmedit/datasets/pipelines/augmentation.py	98.49% <100.00%> (ø)
mmedit/apis/video_interpolation_inference.py	86.40% <0.00%> (-4.83%)	⬇️
mmedit/datasets/builder.py	93.22% <0.00%> (-1.61%)	⬇️
mmedit/apis/restoration_video_inference.py	18.96% <0.00%> (-1.38%)	⬇️
...ls/components/stylegan2/generator_discriminator.py	84.93% <0.00%> (-1.21%)	⬇️
...t/models/video_interpolators/basic_interpolator.py	97.36% <0.00%> (+0.04%)	⬆️
mmedit/datasets/pipelines/blur_kernels.py	84.55% <0.00%> (+1.47%)	⬆️
mmedit/apis/train.py	30.08% <0.00%> (+12.08%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 696e3a8...cd1328d. Read the comment docs.

[wdmwhh]

Because flip is used as a multiplier in line 479 and 480, so -1 stands for flip and 1 stands for no flip. Thus flip should be an 'inverse' flag as the result of the comparison.

Is it correct?

Yes, it is.

[wangruohui]

It seems there is another error. When flipping, the center is shifted.
I made some modifications and added some unit tests, you can have a test to verify.
I am not sure whether it is related to version of opencv

@wdmwhh, if the modification works, please let me know and I will clean the code and merge this PR.

[wdmwhh]

It seems there is another error. When flipping, the center is shifted. I made some modifications and added some unit tests, you can have a test to verify. I am not sure whether it is related to version of opencv

@wdmwhh, if the modification works, please let me know and I will clean the code and merge this PR.

Yes, it makes sense. Thanks for your kindness.