tfa.image.cutout

View source on GitHub

Apply cutout (https://arxiv.org/abs/1708.04552) to images.

This operation applies a (mask_height x mask_width) mask of zeros to a location within images specified by the offset. The pixel values filled in will be of the value replace. The located where the mask will be applied is randomly chosen uniformly over the whole images.

images A tensor of shape (batch_size, height, width, channels) (NHWC), (batch_size, channels, height, width) (NCHW).
mask_size Specifies how big the zero mask that will be generated is that is applied to the images. The mask will be of size (mask_height x mask_width). Note: mask_size should be divisible by 2.
offset A tuple of (height, width) or (batch_size, 2)
constant_values What pixel value to fill in the images in the area that has the cutout mask applied to it.
data_format A string, one of channels_last (default) or channels_first. The ordering of the dimensions in the inputs. channels_last corresponds to inputs with shape (batch_size, ..., channels) while channels_first corresponds to inputs with shape (batch_size, channels, ...).

An image Tensor.

InvalidArgumentError if mask_size can't be divisible by 2.