Use NEAREST NEIGHBOR method when resizing image mask

[balvisio]

IIUC, when resizing the mask, the NEAREST NEIGHBOR interpolation method should be used to prevent the introduction of values of labels that are not present in the original mask.

Currently, if we print the values of sample mask we get:

>>> import numpy as np
>>> mask = iter(train_batches).next()[1]))
>>> print(np.unique(mask))

[0.0000000e+00 1.2207031e-04 3.6621094e-04 ... 1.9995728e+00 1.9999390e+00
 2.0000000e+00]

After the change:

>>> import numpy as np
>>> mask = iter(train_batches).next()[1]))
>>> print(np.unique(mask))

[0. 1. 2.]

[github-actions]

Preview

Preview and run these notebook edits with Google Colab:

• site/en/tutorials/images/segmentation.ipynb

Rendered notebook diffs available on ReviewNB.com.

Format and style

Use the TensorFlow docs notebook tools to format for consistent source diffs and lint for style:

$ python3 -m pip install -U --user git+https://github.com/tensorflow/docs

$ python3 -m tensorflow_docs.tools.nbfmt notebook.ipynb

$ python3 -m tensorflow_docs.tools.nblint --arg=repo:tensorflow/docs notebook.ipynb

If commits are added to the pull request, synchronize your local branch: git pull origin ba/fix-mask-resize-interpolation

[balvisio]

Hello, @MarkDaoust @8bitmp3 would it be possible for someone to review this? Thank you!

[MarkDaoust]

Thanks for the ping.

[balvisio]

The CI check is failing due to formatting issues in another notebook site/en/tutorials/video/transfer_learning_with_movinet.ipynb.