Automated capture of human and object movement

To be able to make computer simulations that are realistic and useful, it is essential to capture real movements. These types of capture are traditionally tedious and time-consuming. In this second lesson, the Pʳ Pai will describe modern imaging and measurement systems, which make modeling made from measurements easier to implement. Medical imaging techniques, in particular fMRI, make it possible to acquire anatomical and behavioral models of human subjects at unprecedented levels of detail. High-speed motion capture systems enable precise measurements of complex human movements such as facial expressions. Finally, new robotic systems capable of examining physical objects enable the acquisition of behavioral models. He will give examples of work in his laboratory to illustrate how this is done. Examples include the UBC Active Measurement Device (ACME), a robotic device for modeling, from real data, physical properties of objects, such as their sound, shape, appearance, response to deformation and surface roughness.