Add support for more shapes in FilterResponseNormalization

[Squadrick]

#765 added FRN to TFA.

Currently, tfa.layers.FilterResponseNormalization only support 4D tensors of order NHWC. We need to add support for arbitrary tensors and let the user specify the axis/axes to normalize over. (axes currently is forced to be [1, 2]).

See the TODO in tensorflow_addons/layers/normalizations.py.

BLOCKED: Till #765 is merged.

[Opletts]

@Squadrick I'll take this up.

[Suraj-Upadhyay]

Hii Guys, I am working on this. Currently reading the research paper for Filter Response Normalization.

If in any doubts I will post it in the comments here and I hope to send a PR soon.

[Opletts]

Sorry guys, I completely forgot about this issue! I have #1496 open and I'll finish it this week.

[seanpmorgan]

TensorFlow Addons is transitioning to a minimal maintenance and release mode. New features will not be added to this repository. For more information, please see our public messaging on this decision:
TensorFlow Addons Wind Down

Please consider sending feature requests / contributions to other repositories in the TF community with a similar charters to TFA:
Keras
Keras-CV
Keras-NLP