The latest research, conducted by Samsung's AI Centre, looked specifically at making a system that can recreate lifelike motion from only one single frame of a person's face. This basically means, using a still image of either a painting or just a normal photograph to make it appear as if it is speaking.
It's not always perfect, but when it works it is pretty amazing. In some examples of the procedure, the famous Mona Lisa can be seen talking and moving like a real person.
Another great paper from Samsung AI lab! @egorzakharovdl et al. animate heads using only few shots of target person… https://t.co/McHEDDVTBo
According to TechCrunch, a lot of data is used to make all of this work. Something called a Generative Adversarial Network is used specifically in the process, which means two models are put against one another, with "one trying to fool the other into thinking what it creates is real".
The process only works on faces and upper torso's so far, but produces some very-realistic looking shots.
If you're interested in the science of this procedure, this is a good explainer video.