Great stuff from the guys and dolls at GOOG – showing the power of AI and machine learning in every day situation, making math fun … doesn’t have to be cat videos all the time, right?
From the paper, titled: “A Neural Representation of Sketch Drawings“, we get this:
We present sketch-rnn, a recurrent neural network (RNN) able to construct stroke-based drawings of common objects. The model is trained on thousands of crude human-drawn images representing hundreds of classes. We outline a framework for conditional and unconditional sketch generation, and describe new robust training methods for generating coherent sketch drawings in a vector format.
And they conclude with:
In this work, we develop a methodology to model sketch drawings using recurrent neural networks. sketch-rnn is able to generate possible ways to finish an existing, but unfinished sketch drawing. Our model can also encode existing sketches into a latent vector, and generate similar looking sketches conditioned on the latent space. We demonstrate what it means to interpolate between two different sketches by interpolating between its latent space. After training on a particular class of image, our
model is able to correct features of an input it deems to be incorrect, and output a similar image with corrected properties. We also show that we can manipulate attributes of a sketch by augmenting the latent space, and demonstrate the importance of enforcing a prior distribution on the latent vector for coherent vector image generation.
Good stuff – and Happy Easter, so go on and draw some bunnies.