Gaussian blur

[artu1999]

Implementation of randomised Gaussian Blur as discussed in #26 and #143, using BaseAugmentationLayer.

Example:

At the moment, only sigma is randomized. vectorized_map doesn't seem to handle kernel of different sizes when distributing depthwise_conv2d between different threads, as when I tried using random size as well I would get this error:

2022-04-01 10:53:14.692102: W tensorflow/core/framework/op_kernel.cc:1745] OP_REQUIRES failed at list_kernels.h:458 : INVALID_ARGUMENT: PartialTensorShape: Incompatible shapes during merge: [3] vs. [2]
Traceback (most recent call last):
File "keras_cv/layers/preprocessing/gaussian_blur_test.py", line 119, in
test.test_on_real_img()
File "keras_cv/layers/preprocessing/gaussian_blur_test.py", line 109, in test_on_real_img
for image in blurred.take(1):
File "/Users/arturosalmi/opt/anaconda3/lib/python3.7/site-packages/tensorflow/python/data/ops/iterator_ops.py", line 766, in next
return self._next_internal()
File "/Users/arturosalmi/opt/anaconda3/lib/python3.7/site-packages/tensorflow/python/data/ops/iterator_ops.py", line 752, in _next_internal
output_shapes=self._flat_output_shapes)
File "/Users/arturosalmi/opt/anaconda3/lib/python3.7/site-packages/tensorflow/python/ops/gen_dataset_ops.py", line 3017, in iterator_get_next
_ops.raise_from_not_ok_status(e, name)
File "/Users/arturosalmi/opt/anaconda3/lib/python3.7/site-packages/tensorflow/python/framework/ops.py", line 7164, in raise_from_not_ok_status
raise core._status_to_exception(e) from None # pylint: disable=protected-access
tensorflow.python.framework.errors_impl.InvalidArgumentError: PartialTensorShape: Incompatible shapes during merge: [3] vs. [2]
[[{{node gaussian_blur/loop_body/range/pfor/TensorListConcatV2}}]] [Op:IteratorGetNext]

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

For more information, open the CLA check for this pull request.

[LukeWood]

Hey @artu1999 thanks for the PR!! Can you accept the CLA? Then we can proceed.