► Keras 3 API documentation / Layers API / Preprocessing layers / Image augmentation layers / MixUp layer

MixUp layer

[source]

MixUp class

keras.layers.MixUp(alpha=0.2, data_format=None, seed=None, **kwargs)

MixUp implements the MixUp data augmentation technique.

Arguments

  • alpha: Float between 0 and 1. Controls the blending strength. Smaller values mean less mixing, while larger values allow for more blending between images. Defaults to 0.2, recommended for ImageNet1k classification.
  • seed: Integer. Used to create a random seed.

References

Example

(images, labels), _ = keras.datasets.cifar10.load_data()
images, labels = images[:8], labels[:8]
labels = keras.ops.cast(keras.ops.one_hot(labels.flatten(), 10), "float32")
mix_up = keras.layers.MixUp(alpha=0.2)
output = mix_up({"images": images, "labels": labels})
MixUp layer
MixUp class