The first 64 images from the training data of CIFAR10 can be plotted using the following code. Doing this, we can get a glimpse of the type of images contained in the dataset:

# Plot of first 64 pictures
par(mfrow = c(8,8), mar = rep(0, 4))
for (i in 1:64) plot(as.raster(trainx[i,,,], max = 255))
par(mfrow = c(1,1))

The images from CIFAR10 are all 32 x 32 color images. The following plot shows 64 images in an 8 x 8 grid:

From the preceding images, we can see that these images come with various backgrounds and are of low resolution. In addition, sometimes, these images aren't completely visible, which makes image classification a challenging task.