ยป Keras API reference / KerasCV / Layers / Preprocessing layers / FourierMix layer

FourierMix layer

[source]

FourierMix class

keras_cv.layers.FourierMix(alpha=0.5, decay_power=3, seed=None, **kwargs)

FourierMix implements the FMix data augmentation technique.

Arguments

  • alpha: Float value for beta distribution. Inverse scale parameter for the gamma distribution. This controls the shape of the distribution from which the smoothing values are sampled. Defaults to 0.5, which is a recommended value in the paper.
  • decay_power: A float value representing the decay power. Defaults to 3, as recommended in the paper.
  • seed: Integer. Used to create a random seed.

References

Sample usage:

(images, labels), _ = tf.keras.datasets.cifar10.load_data()
fourier_mix = keras_cv.layers.preprocessing.FourierMix(0.5)
augmented_images, updated_labels = fourier_mix({'images': images, 'labels': labels})
# output == {'images': updated_images, 'labels': updated_labels}