So you’ve taken what would be the perfect group picture, except that Uncle Roy is staring at your sister-in-law’s bosom. A new technology called DeepWarp can change that part without making Roy look unnatural.
In this work, we consider the task of generating highly-realistic images of a given face with a redirected gaze. We treat this problem as a specific instance of conditional image generation, and suggest a new deep architecture that can handle this task very well as revealed by numerical comparison with prior art and a user study. Our deep architecture performs coarse-to-fine warping with an additional intensity correction of individual pixels. All these operations are performed in a feed-forward manner, and the parameters associated with different operations are learned jointly in the end-to-end fashion. After learning, the resulting neural network can synthesize images with manipulated gaze, while the redirection angle can be selected arbitrarily from a certain range and provided as an input to the network.
I didn’t understand any of that, but that’s not what’s important. What’s really interesting are the examples given on the project page. You can pull up any of 16 pictures of faces and move their eyes around. They can look you up and down, they can shift left and right right, or they can roll around. It’s a shame they can’t move independently of each other, but what you will see is creepy and funny enough as it is. -via Metafilter