FR for keras.preprocessing.image_dataset_from_directory label_mode parameter to accept enum as input

[BlueskyFR]

System information

• TensorFlow version (you are using): 2.4.1
• Are you willing to contribute it (Yes/No): Yes

Describe the feature and the current behavior/state.

When calling the tf.keras.preprocessing.image_dataset_from_directory() function, the parameter label_mode= accepts 3 values: int, categorical and binary. However, the string object is not deterministic: for that, I propose to add another option, which would be to specify the value using an enumeration which IS deterministic.

Will this change the current api? How?

No

Who will benefit with this feature?

Users who want more structured code.

Any Other info.

This behavior is, as far as I know, already available for some features in Keras.

[fchollet]

However, the string object is not deterministic: for that, I propose to add another option, which would be to specify the value using an enumeration which IS deterministic.

What does that mean?

[BlueskyFR]

However, the string object is not deterministic: for that, I propose to add another option, which would be to specify the value using an enumeration which IS deterministic.

What does that mean?

By accepting a string as input, typos can be made and no autocompletion can be offered by the IDE, which is however the case when using an enum.

My proposition is to add support for an alternative input type using an enumeration, such as tf.keras.preprocessing.label_mode.[INT|CATEGORICAL|BINARY] so that autocompletion can be made, typos are avoided, and users would still be free to specify a string "categorical" instead of tf.keras.preprocessing.label_mode.CATEGORICAL if they want.

What do you think?

[sachinprasadhs]

tf.keras.preprocessing.image_dataset_from_directory will be deprecated in future Tensorflow version, prefer loading data with tf.keras.utils.image_dataset_from_directory, and then transforming the output tf.data.Dataset with preprocessing layers.
Let us know if you still face an issue. Thanks!

[google-ml-butler]

This issue has been automatically marked as stale because it has no recent activity. It will be closed if no further activity occurs. Thank you.

[google-ml-butler]

Closing as stale. Please reopen if you'd like to work on this further.