fix compute_output_shape behavior in normalizations.py

[Zhaopudark]

Description

Fixes #2677

Normalization layers in normalizations.py have been write very well. Thanks for the contributors. These layers work well in most cases, but their behavor in compute_output_shape is very smallly different from their description and different from tf.keras.layers.Layer's design ideas about compute_output_shape .

Since tfa.layers.GroupNormalization, tfa.layers.InstanceNormalization and tfa.layers.FilterResponseNormalization are all inherited from tf.keras.layers.Layer, they all override compute_output_shape function.

However, there is no need to override this function, because tf.keras.layers.Layer has done it well, i.e., if call and build functions have been overrided, the layer can automaticly give out correct output_shape buy compute_output_shape in superclass. Remove the overrided compute_output_shape, the bug will disappear.

Actually, overriding compute_output_shape when inherite a custom layer from tf.keras.layer.Layer is usually unnecessary in most cases, unless a user wants an output_shape that different from call's output. But in normalizations.py, normalization ops do not change shapes, the special above case is not suitable.

Type of change

• Bug fix
• Additional Testing

Checklist:

• I've properly formatted my code according to the guidelines
  • By running Black + Flake8
  • By running pre-commit hooks
• This PR addresses an already submitted issue for TensorFlow Addons
• I have added tests that prove my fix is effective or that my feature works

[bot-of-gabrieldemarmiesse]

@Smokrow

You are owner of some files modified in this pull request.
Would you kindly review the changes whenever you have the time to?
Thank you very much.