Skip to contents

An image_tensor() object that is a super thin wrapper around a 4D base::array(), representing image data in the format "b h w c" (batch, height, width, channels). The actual image data is 6 black letters on differently colored backgrounds, spelling out the word "einops". When printing this object in the terminal it will automatically plot the images. To subset, use the [ operator, and when used with a single index, it will return a single image tensor.

Usage

einops_image

Format

An image_tensor() object with 6 images, each of size 96 x 96 pixels, with 3 color channels (RGB). The images are stored in a 4D array with dimensions (6, 96, 96, 3).

Examples

data("einops_image")
for (i in seq_len(6)) print(einops_image[i, , , ])