tfm.vision.preprocess_ops.random_jpeg_quality

Randomly encode the image as jpeg and decode it.

tfm.vision.preprocess_ops.random_jpeg_quality(
    image: tf.Tensor,
    min_quality: (int | tf.Tensor) = 20,
    max_quality: (int | tf.Tensor) = 100,
    prob_to_apply: (float | tf.Tensor) = 0.6
) -> tf.Tensor

Args
`image`	a uint8 'Tensor' of shape [height, width, 3] representing the input image.
`min_quality`	minimum jpeg quality in range of [0, 100].
`max_quality`	maximum jpeg quality in range of [0, 100].
`prob_to_apply`	probability to apply this augmentation.

Returns
image with jpeg quality changed

Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates. Some content is licensed under the numpy license.

Last updated 2026-11-06 UTC.

tfm.vision.preprocess_ops.random_jpeg_quality Stay organized with collections Save and categorize content based on your preferences.

Args

Returns

tfm.vision.preprocess_ops.random_jpeg_quality