Image Recognition
Fashion MNIST
Classify grayscale images of fashion items
fashion

This is a dataset consisting of 28x28 pixels, grayscale images taken from Zalando catalog, distributed among 10 classes.

The training set is composed of 60k images and the test set 10k images.

The characteristics of this dataset (28x28, BnW, number of classes etc) are the same as MNIST on purpose. Zalando Research intends for this this dataset to be a drop-in replacement for it.

Here are the classes:

  • 0 == T-shirt/Top
  • 1 == Trouser
  • 2 == Pullover
  • 3 == Dress
  • 4 == Coat
  • 5 == Sandal
  • 6 == Shirt
  • 7 == Sneaker
  • 8 == Bag
  • 9 == Ankle boot
Results from deployment:
85%
5.1kB
Accuracy
RAM used