Computers might become the artists of tomorrow. Professor Ruslan Salakhutdinov and his research team might make that happen. Salakhutdinov was highlighted in Bloomberg news for his work on a paper that introduces a model that draws images from text descriptions.
What is significant about this model is that they were able to train a neural network system to draw images using text as the input, as opposed to other models that can take images and describe their contents.
Professor Salakhutdinov and his research team Elman Mansimov (HBSc 2015), Emilio Parisotto (PhD, CompSci) and Jimmy Lei Ba (ECE), were recently interviewed by Vice’s online technology magazine Motherboard concerning this same research.
Read more at Bloomberg news