How to take some sub-directories from a main directory tf.keras.utils.image_dataset_from_directory

[ARPIT2128]

How to take some sub-directories from a main directory

• Hello I'm new to this issue stuff 😃
• Basically I was unable to find a fix for the issue on Stackoverflow so I'm here to ask the experts. 💜
• Hope I'm not creating a problem for you all will make it short and simple. 😥

def image_dataset_from_directory(
    directory,
    labels="inferred",
    label_mode="int",
    class_names=None,
    color_mode="rgb",
    batch_size=32,
    image_size=(256, 256),
    shuffle=True,
    seed=None,
    validation_split=None,
    subset=None,
    interpolation="bilinear",
    follow_links=False,
    crop_to_aspect_ratio=False,
    **kwargs,
):

• HERE IF I ONLY WANT class_a THEN WHAT TO DO

    main_directory/
    ...class_a/
    ......a_image_1.jpg
    ......a_image_2.jpg
    ...class_b/
    ......b_image_1.jpg
    ......b_image_2.jpg

• As shown in above figure I'm trying to use the tf.keras.utils.image_dataset_from_directory
• But here I want to take some sub directories only and not all like .git files from the main directory
• Previously it was in from tensorflow.keras.preprocessing.image import image_dataset_from_directory
• with parameter ignore
• is there a easy fix for this using this directory or any relevant.
  THANKS IN ADVANCE 😉

[sachinprasadhs]

Hi,

Thanks for reporting the issue.
I was able to verify the issue in colab.

I have created a PR to handle the hidden folders like .git , .ipynb etc.

Coming to your question about ignoring directory with directory name as input is not possible in the current implementation as per my understanding.

[google-ml-butler]

Are you satisfied with the resolution of your issue?
Yes
No