Draw a character below and see an autoencoder reconstruct it.
Draw a character in the black box below and see an autoencoder reconstruct it. Left click uses white, right click uses black but you can adjust the color of the pens. The top right grey box will contain the best guess of the neural network. The middle right grey box will contain the reconstructed image of your drawing using a convolutional autoencoder. The bottom right grey box will contain the reconstructed image of your drawing using a variational autoencoder .