Digital Domain are doing great work on performance capture. All of these technique will trickle down to 3D artist and animators in the future.
My main interest in this is their use of machine learning. They are using it to discover form and sort through optical flow. From my own research I know that AI really does have the potential to create digital actors, with their own identifying characteristics. We will see, in the future, 3d artists not only producing visual concepts for characters but also designing the physical movements that define their characters. We will need to not only be thinking about anatomy but about how our physical movement express internal idea and beliefs.
I hope they hurry up their development. I for one can’t wait.